From haehn at bwh.harvard.edu  Thu Jun  2 16:57:36 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Thu, 2 Jun 2011 12:57:36 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText or
 .setButtonBoxHints in Python
Message-ID: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>

Hi guys,

I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
but I can not access the method:

a = ctk.ctkWorkflowWidgetStep()
...
a.setButtonBoxHints(..) # does not exist

Accessing the ButtonBoxHint enum with its values is possible.

Also, I am missing access to a.setBackButtonText(..).

Am I doing something wrong or is the doxygen documentation outdated?

Thanks,
Daniel


From pieper at ibility.net  Thu Jun  2 17:06:57 2011
From: pieper at ibility.net (Steve Pieper)
Date: Thu, 2 Jun 2011 13:06:57 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
Message-ID: <BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>

Hi Daniel -

With the PythonQt wrapping, only properties, signals/slots, and
methods tagged with Q_INVOKABLE are available in python.

I would argue that we should try to expose the public API of our
widgets as much as possible to python but right now there is a lot of
functionality that is only available via C++.

-Steve

On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
> Hi guys,
>
> I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
> but I can not access the method:
>
> a = ctk.ctkWorkflowWidgetStep()
> ...
> a.setButtonBoxHints(..) # does not exist
>
> Accessing the ButtonBoxHint enum with its values is possible.
>
> Also, I am missing access to a.setBackButtonText(..).
>
> Am I doing something wrong or is the doxygen documentation outdated?
>
> Thanks,
> Daniel
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>


From jchris.fillionr at kitware.com  Thu Jun  2 17:07:15 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Thu, 2 Jun 2011 13:07:15 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
Message-ID: <BANLkTimFk2-V1R36PPqByRJFhkeiFv1euQ@mail.gmail.com>

Hi Daniel,

Please consider the following remarks,

* The doxygen documentation is up-to-date. See
http://www.commontk.org/docs/html/index.html
The text on the bottom right corner indicates: june 2nd

* The doxygen documentation refers to the Cpp API. As of today, there are no
indication reported letting the user know if the method is available from
python.

* Looking at the code of
ctkWorkflowWidgetStep<https://github.com/commontk/CTK/blob/master/Libs/Widgets/ctkWorkflowWidgetStep.h>,
backButtonText is added to the meta object system as a property (using
Q_PROPERTY). It means, it can be set/get using the following syntax:

a.backButtonText = "Play again"
print a.backButtonText

Hth
Jc

On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi guys,
>
> I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
> but I can not access the method:
>
> a = ctk.ctkWorkflowWidgetStep()
> ...
> a.setButtonBoxHints(..) # does not exist
>
> Accessing the ButtonBoxHint enum with its values is possible.
>
> Also, I am missing access to a.setBackButtonText(..).
>
> Am I doing something wrong or is the doxygen documentation outdated?
>
> Thanks,
> Daniel
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110602/17226c74/attachment.htm>

From jchris.fillionr at kitware.com  Thu Jun  2 18:26:33 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Thu, 2 Jun 2011 14:26:33 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
	<BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>
Message-ID: <BANLkTinZNpasf8q=B-ZTXG=XzvdAFQrcdA@mail.gmail.com>

On Thu, Jun 2, 2011 at 1:06 PM, Steve Pieper <pieper at ibility.net> wrote:

> Hi Daniel -
>
> With the PythonQt wrapping, only properties, signals/slots, and
> methods tagged with Q_INVOKABLE are available in python.
>

Correct.


> I would argue that we should try to expose the public API of our
> widgets as much as possible to python but right now there is a lot of
> functionality that is only available via C++.
>

When the API of class is designed, efforts are put to elaborate it.
Similarly, the same thought process should be done regarding the usability
of the object from within python.

To validate the API, a set of test *are* (should be) developed. As of today,
these tests are written in Cpp. We could think of having tests also written
in Python.

  -

>
> -Steve
>
> On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
> wrote:
> > Hi guys,
> >
> > I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
> > but I can not access the method:
> >
> > a = ctk.ctkWorkflowWidgetStep()
> > ...
> > a.setButtonBoxHints(..) # does not exist
> >
> > Accessing the ButtonBoxHint enum with its values is possible.
> >
> > Also, I am missing access to a.setBackButtonText(..).
> >
> > Am I doing something wrong or is the doxygen documentation outdated?
> >
> > Thanks,
> > Daniel
> > _______________________________________________
> > Ctk-developers mailing list
> > Ctk-developers at commontk.org
> > http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110602/84d2de0b/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 00:20:35 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Thu, 2 Jun 2011 20:20:35 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTinZNpasf8q=B-ZTXG=XzvdAFQrcdA@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
	<BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>
	<BANLkTinZNpasf8q=B-ZTXG=XzvdAFQrcdA@mail.gmail.com>
Message-ID: <BANLkTi=bLU3RcMZqLOMoKucT50Jd46nCWg@mail.gmail.com>

Steve, JC, thanks for the detailed replies!

Is there something which would prevent making the buttonhints for the wizard
steps Q_INVOKABLE?

Cheers,
Daniel
On Jun 2, 2011 2:26 PM, "Jean-Christophe Fillion-Robin" <
jchris.fillionr at kitware.com> wrote:
> On Thu, Jun 2, 2011 at 1:06 PM, Steve Pieper <pieper at ibility.net> wrote:
>
>> Hi Daniel -
>>
>> With the PythonQt wrapping, only properties, signals/slots, and
>> methods tagged with Q_INVOKABLE are available in python.
>>
>
> Correct.
>
>
>> I would argue that we should try to expose the public API of our
>> widgets as much as possible to python but right now there is a lot of
>> functionality that is only available via C++.
>>
>
> When the API of class is designed, efforts are put to elaborate it.
> Similarly, the same thought process should be done regarding the usability
> of the object from within python.
>
> To validate the API, a set of test *are* (should be) developed. As of
today,
> these tests are written in Cpp. We could think of having tests also
written
> in Python.
>
> -
>
>>
>> -Steve
>>
>> On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> wrote:
>> > Hi guys,
>> >
>> > I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
>> > but I can not access the method:
>> >
>> > a = ctk.ctkWorkflowWidgetStep()
>> > ...
>> > a.setButtonBoxHints(..) # does not exist
>> >
>> > Accessing the ButtonBoxHint enum with its values is possible.
>> >
>> > Also, I am missing access to a.setBackButtonText(..).
>> >
>> > Am I doing something wrong or is the doxygen documentation outdated?
>> >
>> > Thanks,
>> > Daniel
>> > _______________________________________________
>> > Ctk-developers mailing list
>> > Ctk-developers at commontk.org
>> > http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>
>
>
> --
> +1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110602/fe3b72fe/attachment.htm>

From jchris.fillionr at kitware.com  Fri Jun  3 03:46:49 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Thu, 2 Jun 2011 23:46:49 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTi=bLU3RcMZqLOMoKucT50Jd46nCWg@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
	<BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>
	<BANLkTinZNpasf8q=B-ZTXG=XzvdAFQrcdA@mail.gmail.com>
	<BANLkTi=bLU3RcMZqLOMoKucT50Jd46nCWg@mail.gmail.com>
Message-ID: <BANLkTik3G7Qb-dna5i1s=ANL_QW+V5S3rA@mail.gmail.com>

Hi Daniel,

The requested feature has been implemented, see
https://github.com/commontk/CTK/commit/e8b169829d679e07fd13d2878960af0784edaed8

Please refer to the comment for more details.

Thanks
Jc

On Thu, Jun 2, 2011 at 8:20 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Steve, JC, thanks for the detailed replies!
>
> Is there something which would prevent making the buttonhints for the
> wizard steps Q_INVOKABLE?
>
> Cheers,
> Daniel
> On Jun 2, 2011 2:26 PM, "Jean-Christophe Fillion-Robin" <
> jchris.fillionr at kitware.com> wrote:
> > On Thu, Jun 2, 2011 at 1:06 PM, Steve Pieper <pieper at ibility.net> wrote:
> >
> >> Hi Daniel -
> >>
> >> With the PythonQt wrapping, only properties, signals/slots, and
> >> methods tagged with Q_INVOKABLE are available in python.
> >>
> >
> > Correct.
> >
> >
> >> I would argue that we should try to expose the public API of our
> >> widgets as much as possible to python but right now there is a lot of
> >> functionality that is only available via C++.
> >>
> >
> > When the API of class is designed, efforts are put to elaborate it.
> > Similarly, the same thought process should be done regarding the
> usability
> > of the object from within python.
> >
> > To validate the API, a set of test *are* (should be) developed. As of
> today,
> > these tests are written in Cpp. We could think of having tests also
> written
> > in Python.
> >
> > -
> >
> >>
> >> -Steve
> >>
> >> On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
> >> wrote:
> >> > Hi guys,
> >> >
> >> > I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in Python
> >> > but I can not access the method:
> >> >
> >> > a = ctk.ctkWorkflowWidgetStep()
> >> > ...
> >> > a.setButtonBoxHints(..) # does not exist
> >> >
> >> > Accessing the ButtonBoxHint enum with its values is possible.
> >> >
> >> > Also, I am missing access to a.setBackButtonText(..).
> >> >
> >> > Am I doing something wrong or is the doxygen documentation outdated?
> >> >
> >> > Thanks,
> >> > Daniel
> >> > _______________________________________________
> >> > Ctk-developers mailing list
> >> > Ctk-developers at commontk.org
> >> > http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >> >
> >> _______________________________________________
> >> Ctk-developers mailing list
> >> Ctk-developers at commontk.org
> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >>
> >
> >
> >
> > --
> > +1 919 869 8849
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110602/05e619eb/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 11:37:04 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 07:37:04 -0400
Subject: [Ctk-developers] Using ctkWorkflowWidgetStep.setBackButtonText
 or .setButtonBoxHints in Python
In-Reply-To: <BANLkTik3G7Qb-dna5i1s=ANL_QW+V5S3rA@mail.gmail.com>
References: <BANLkTi=iz+g2ZcMYQFGJqmn2SRDCjqocug@mail.gmail.com>
	<BANLkTi=KM-+4Cfzt56+0B44ORxgcD4uLqg@mail.gmail.com>
	<BANLkTinZNpasf8q=B-ZTXG=XzvdAFQrcdA@mail.gmail.com>
	<BANLkTi=bLU3RcMZqLOMoKucT50Jd46nCWg@mail.gmail.com>
	<BANLkTik3G7Qb-dna5i1s=ANL_QW+V5S3rA@mail.gmail.com>
Message-ID: <BANLkTinWMnFKhbTCjD7jSxjUsHXpjXy_YQ@mail.gmail.com>

Great!

Thanks a lot,
Daniel
On Jun 2, 2011 11:47 PM, "Jean-Christophe Fillion-Robin" <
jchris.fillionr at kitware.com> wrote:
> Hi Daniel,
>
> The requested feature has been implemented, see
>
https://github.com/commontk/CTK/commit/e8b169829d679e07fd13d2878960af0784edaed8
>
> Please refer to the comment for more details.
>
> Thanks
> Jc
>
> On Thu, Jun 2, 2011 at 8:20 PM, Daniel Haehn <haehn at bwh.harvard.edu>
wrote:
>
>> Steve, JC, thanks for the detailed replies!
>>
>> Is there something which would prevent making the buttonhints for the
>> wizard steps Q_INVOKABLE?
>>
>> Cheers,
>> Daniel
>> On Jun 2, 2011 2:26 PM, "Jean-Christophe Fillion-Robin" <
>> jchris.fillionr at kitware.com> wrote:
>> > On Thu, Jun 2, 2011 at 1:06 PM, Steve Pieper <pieper at ibility.net>
wrote:
>> >
>> >> Hi Daniel -
>> >>
>> >> With the PythonQt wrapping, only properties, signals/slots, and
>> >> methods tagged with Q_INVOKABLE are available in python.
>> >>
>> >
>> > Correct.
>> >
>> >
>> >> I would argue that we should try to expose the public API of our
>> >> widgets as much as possible to python but right now there is a lot of
>> >> functionality that is only available via C++.
>> >>
>> >
>> > When the API of class is designed, efforts are put to elaborate it.
>> > Similarly, the same thought process should be done regarding the
>> usability
>> > of the object from within python.
>> >
>> > To validate the API, a set of test *are* (should be) developed. As of
>> today,
>> > these tests are written in Cpp. We could think of having tests also
>> written
>> > in Python.
>> >
>> > -
>> >
>> >>
>> >> -Steve
>> >>
>> >> On Thu, Jun 2, 2011 at 12:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> >> wrote:
>> >> > Hi guys,
>> >> >
>> >> > I am trying to hide the buttonbox on a ctkWorkflowWidgetStep in
Python
>> >> > but I can not access the method:
>> >> >
>> >> > a = ctk.ctkWorkflowWidgetStep()
>> >> > ...
>> >> > a.setButtonBoxHints(..) # does not exist
>> >> >
>> >> > Accessing the ButtonBoxHint enum with its values is possible.
>> >> >
>> >> > Also, I am missing access to a.setBackButtonText(..).
>> >> >
>> >> > Am I doing something wrong or is the doxygen documentation outdated?
>> >> >
>> >> > Thanks,
>> >> > Daniel
>> >> > _______________________________________________
>> >> > Ctk-developers mailing list
>> >> > Ctk-developers at commontk.org
>> >> > http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >> >
>> >> _______________________________________________
>> >> Ctk-developers mailing list
>> >> Ctk-developers at commontk.org
>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >>
>> >
>> >
>> >
>> > --
>> > +1 919 869 8849
>>
>
>
>
> --
> +1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/d72683e1/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 16:17:23 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 12:17:23 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
Message-ID: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>

Hi devels,

in C++ I can do

  this->setName("Title of step");
  this->setDescription("Instructions");

in a ctkWorkflowWidgetStep, which then get nicely displayed in the
associated ctkWorkflowGroupBox.

In Python, the same methods can be accessed but the
ctkWorkflowGroupBox does not seem to show them.

Do I need to trigger an update on the box or something similar?

Thank you!
Daniel


From haehn at bwh.harvard.edu  Fri Jun  3 19:13:39 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 15:13:39 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
Message-ID: <BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>

Hi guys,

I got it:

the ctkWorkflowGroupBox listens for the currentStepChanged signal to
update the title (==name) and subTitle (==description) for the current
step.

So maybe it makes sense to invoke this signal when the
ctkWorkflow.start() gets called?

If I do it manually like that, it works:

      self.workflow.start()
      self.workflow.currentStepChanged(steps[0])

Cheers,
Daniel

On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
> Hi devels,
>
> in C++ I can do
>
> ?this->setName("Title of step");
> ?this->setDescription("Instructions");
>
> in a ctkWorkflowWidgetStep, which then get nicely displayed in the
> associated ctkWorkflowGroupBox.
>
> In Python, the same methods can be accessed but the
> ctkWorkflowGroupBox does not seem to show them.
>
> Do I need to trigger an update on the box or something similar?
>
> Thank you!
> Daniel
>


From haehn at bwh.harvard.edu  Fri Jun  3 19:34:51 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 15:34:51 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
Message-ID: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>

Hi guys,

if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
along the workflow (according to onEntry, validate and onExit calls).

But then, after reaching step x, it goes directly back to where I am
coming from - again along the workflow. This goes so fast so it seems
there is no movement at all.

I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
right? After reaching the 'finish step' (which in my understanding is
the target step), it goes back to the starting step.


1019 // --------------------------------------------------------------------------
1020 void ctkWorkflow::goToStepSucceeded()
1021 {
1022   Q_D(ctkWorkflow);
1023
1024   logger.debug("goToStepSucceeded");
1025
1026   // after success, go back to the step at which we begin looking for
1027   // the finish step (will exit the current step and enter the
starting step)
1028
1029   d->createTransitionToPreviousStartingStep(d->StartingStep,
d->CurrentStep);
1030
1031   d->GoToStep = 0;
1032   d->StartingStep->setStatusText("Attempt to go to the finish
step succeeded");
1033   d->StartingStep = 0;
1034
1035   this->goFromGoToStepToStartingStep();
1036 }
1037

Is this correct behavior and I understand the goToStep(x) call wrong
or is this a bug?

Cheers,
Daniel


From danielle.pace at kitware.com  Fri Jun  3 19:40:04 2011
From: danielle.pace at kitware.com (Danielle Pace)
Date: Fri, 3 Jun 2011 15:40:04 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
Message-ID: <BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>

Hi Daniel,

This is the intended behavior when using 'goToStep'.

The idea is that, partway through the workflow, you could execute the steps
until a 'finish step' was reached - using the default parameters for those
steps.  Then you can evaluate the results that occur with default
parameters.  If you get to the 'finish' step successfully, it puts you back
where you started - so that you can continue to step through and alter
parameters if you like.

If you do not get to the 'finish' step successfully (i.e. default parameters
produce errors on your data or because of the parameters you set in prior
steps), it will stop at the error.

Hope that helps,

Danielle





On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi guys,
>
> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
> along the workflow (according to onEntry, validate and onExit calls).
>
> But then, after reaching step x, it goes directly back to where I am
> coming from - again along the workflow. This goes so fast so it seems
> there is no movement at all.
>
> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
> right? After reaching the 'finish step' (which in my understanding is
> the target step), it goes back to the starting step.
>
>
> 1019 //
> --------------------------------------------------------------------------
> 1020 void ctkWorkflow::goToStepSucceeded()
> 1021 {
> 1022   Q_D(ctkWorkflow);
> 1023
> 1024   logger.debug("goToStepSucceeded");
> 1025
> 1026   // after success, go back to the step at which we begin looking for
> 1027   // the finish step (will exit the current step and enter the
> starting step)
> 1028
> 1029   d->createTransitionToPreviousStartingStep(d->StartingStep,
> d->CurrentStep);
> 1030
> 1031   d->GoToStep = 0;
> 1032   d->StartingStep->setStatusText("Attempt to go to the finish
> step succeeded");
> 1033   d->StartingStep = 0;
> 1034
> 1035   this->goFromGoToStepToStartingStep();
> 1036 }
> 1037
>
> Is this correct behavior and I understand the goToStep(x) call wrong
> or is this a bug?
>
> Cheers,
> Daniel
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
Danielle Pace, M.ESc.
Research and Development Engineer
Kitware Inc.,
North Carolina Office

www.kitware.com
919-969-6990 X 319
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/3f453b3d/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 19:45:57 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 15:45:57 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
Message-ID: <BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>

Hi Danielle,

hm this does not make totally sense to me. What about the canGoToStep
method - this only checks for next step?

Wouldn't it make more sense to have the behavior I observed and you
described for the canGoToStep call and do not go back when you call
goToStep?

Or is there another method I could use to go from step 1 to step 4, f.e.?

Thanks
Daniel

On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
> Hi Daniel,
> This is the intended behavior when using 'goToStep'.
> The idea is that, partway through the workflow, you could execute the steps
> until a 'finish step' was reached - using the default parameters for those
> steps. ?Then you can evaluate the results that occur with default
> parameters. ?If you get to the 'finish' step successfully, it puts you back
> where you started - so that you can continue to step through and alter
> parameters if you like.
> If you do not get to the 'finish' step successfully (i.e. default parameters
> produce errors on your data or because of the parameters you set in prior
> steps), it will stop at the error.
> Hope that helps,
>
> Danielle
>
>
>
>
> On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
>>
>> Hi guys,
>>
>> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>> along the workflow (according to onEntry, validate and onExit calls).
>>
>> But then, after reaching step x, it goes directly back to where I am
>> coming from - again along the workflow. This goes so fast so it seems
>> there is no movement at all.
>>
>> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>> right? After reaching the 'finish step' (which in my understanding is
>> the target step), it goes back to the starting step.
>>
>>
>> 1019 //
>> --------------------------------------------------------------------------
>> 1020 void ctkWorkflow::goToStepSucceeded()
>> 1021 {
>> 1022 ? Q_D(ctkWorkflow);
>> 1023
>> 1024 ? logger.debug("goToStepSucceeded");
>> 1025
>> 1026 ? // after success, go back to the step at which we begin looking for
>> 1027 ? // the finish step (will exit the current step and enter the
>> starting step)
>> 1028
>> 1029 ? d->createTransitionToPreviousStartingStep(d->StartingStep,
>> d->CurrentStep);
>> 1030
>> 1031 ? d->GoToStep = 0;
>> 1032 ? d->StartingStep->setStatusText("Attempt to go to the finish
>> step succeeded");
>> 1033 ? d->StartingStep = 0;
>> 1034
>> 1035 ? this->goFromGoToStepToStartingStep();
>> 1036 }
>> 1037
>>
>> Is this correct behavior and I understand the goToStep(x) call wrong
>> or is this a bug?
>>
>> Cheers,
>> Daniel
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
> www.kitware.com
> 919-969-6990 X 319
>


From danielle.pace at kitware.com  Fri Jun  3 19:52:51 2011
From: danielle.pace at kitware.com (Danielle Pace)
Date: Fri, 3 Jun 2011 15:52:51 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
Message-ID: <BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>

Hi Daniel,

CanGoToStep checks to see if a path exists between two steps - so it is not
helpful in your case.

Right now, there is no method to go to a specific step without transitioning
back.  We could add a property that dictates whether to go back to the
starting step after successfully getting to a 'finish' step...

Thanks,

Danielle





On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi Danielle,
>
> hm this does not make totally sense to me. What about the canGoToStep
> method - this only checks for next step?
>
> Wouldn't it make more sense to have the behavior I observed and you
> described for the canGoToStep call and do not go back when you call
> goToStep?
>
> Or is there another method I could use to go from step 1 to step 4, f.e.?
>
> Thanks
> Daniel
>
> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace <danielle.pace at kitware.com>
> wrote:
> > Hi Daniel,
> > This is the intended behavior when using 'goToStep'.
> > The idea is that, partway through the workflow, you could execute the
> steps
> > until a 'finish step' was reached - using the default parameters for
> those
> > steps.  Then you can evaluate the results that occur with default
> > parameters.  If you get to the 'finish' step successfully, it puts you
> back
> > where you started - so that you can continue to step through and alter
> > parameters if you like.
> > If you do not get to the 'finish' step successfully (i.e. default
> parameters
> > produce errors on your data or because of the parameters you set in prior
> > steps), it will stop at the error.
> > Hope that helps,
> >
> > Danielle
> >
> >
> >
> >
> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
> wrote:
> >>
> >> Hi guys,
> >>
> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
> >> along the workflow (according to onEntry, validate and onExit calls).
> >>
> >> But then, after reaching step x, it goes directly back to where I am
> >> coming from - again along the workflow. This goes so fast so it seems
> >> there is no movement at all.
> >>
> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
> >> right? After reaching the 'finish step' (which in my understanding is
> >> the target step), it goes back to the starting step.
> >>
> >>
> >> 1019 //
> >>
> --------------------------------------------------------------------------
> >> 1020 void ctkWorkflow::goToStepSucceeded()
> >> 1021 {
> >> 1022   Q_D(ctkWorkflow);
> >> 1023
> >> 1024   logger.debug("goToStepSucceeded");
> >> 1025
> >> 1026   // after success, go back to the step at which we begin looking
> for
> >> 1027   // the finish step (will exit the current step and enter the
> >> starting step)
> >> 1028
> >> 1029   d->createTransitionToPreviousStartingStep(d->StartingStep,
> >> d->CurrentStep);
> >> 1030
> >> 1031   d->GoToStep = 0;
> >> 1032   d->StartingStep->setStatusText("Attempt to go to the finish
> >> step succeeded");
> >> 1033   d->StartingStep = 0;
> >> 1034
> >> 1035   this->goFromGoToStepToStartingStep();
> >> 1036 }
> >> 1037
> >>
> >> Is this correct behavior and I understand the goToStep(x) call wrong
> >> or is this a bug?
> >>
> >> Cheers,
> >> Daniel
> >> _______________________________________________
> >> Ctk-developers mailing list
> >> Ctk-developers at commontk.org
> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >
> >
> >
> > --
> > Danielle Pace, M.ESc.
> > Research and Development Engineer
> > Kitware Inc.,
> > North Carolina Office
> > www.kitware.com
> > 919-969-6990 X 319
> >
>



-- 
Danielle Pace, M.ESc.
Research and Development Engineer
Kitware Inc.,
North Carolina Office

www.kitware.com
919-969-6990 X 319
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/9b1b486b/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 19:56:46 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 15:56:46 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
	<BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
Message-ID: <BANLkTimL=VysUtYhkir=89ciy6s3krUYHw@mail.gmail.com>

Hi Danielle,

the property sounds like a great idea!

Thanks,
Daniel

On Fri, Jun 3, 2011 at 3:52 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
> Hi Daniel,
> CanGoToStep checks to see if a path exists between two steps - so it is not
> helpful in your case.
> Right now, there is no method to go to a specific step without transitioning
> back. ?We could add a property that dictates whether to go back to the
> starting step after successfully getting to a 'finish' step...
> Thanks,
>
> Danielle
>
>
>
>
> On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
>>
>> Hi Danielle,
>>
>> hm this does not make totally sense to me. What about the canGoToStep
>> method - this only checks for next step?
>>
>> Wouldn't it make more sense to have the behavior I observed and you
>> described for the canGoToStep call and do not go back when you call
>> goToStep?
>>
>> Or is there another method I could use to go from step 1 to step 4, f.e.?
>>
>> Thanks
>> Daniel
>>
>> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace <danielle.pace at kitware.com>
>> wrote:
>> > Hi Daniel,
>> > This is the intended behavior when using 'goToStep'.
>> > The idea is that, partway through the workflow, you could execute the
>> > steps
>> > until a 'finish step' was reached - using the default parameters for
>> > those
>> > steps. ?Then you can evaluate the results that occur with default
>> > parameters. ?If you get to the 'finish' step successfully, it puts you
>> > back
>> > where you started - so that you can continue to step through and alter
>> > parameters if you like.
>> > If you do not get to the 'finish' step successfully (i.e. default
>> > parameters
>> > produce errors on your data or because of the parameters you set in
>> > prior
>> > steps), it will stop at the error.
>> > Hope that helps,
>> >
>> > Danielle
>> >
>> >
>> >
>> >
>> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> > wrote:
>> >>
>> >> Hi guys,
>> >>
>> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>> >> along the workflow (according to onEntry, validate and onExit calls).
>> >>
>> >> But then, after reaching step x, it goes directly back to where I am
>> >> coming from - again along the workflow. This goes so fast so it seems
>> >> there is no movement at all.
>> >>
>> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>> >> right? After reaching the 'finish step' (which in my understanding is
>> >> the target step), it goes back to the starting step.
>> >>
>> >>
>> >> 1019 //
>> >>
>> >> --------------------------------------------------------------------------
>> >> 1020 void ctkWorkflow::goToStepSucceeded()
>> >> 1021 {
>> >> 1022 ? Q_D(ctkWorkflow);
>> >> 1023
>> >> 1024 ? logger.debug("goToStepSucceeded");
>> >> 1025
>> >> 1026 ? // after success, go back to the step at which we begin looking
>> >> for
>> >> 1027 ? // the finish step (will exit the current step and enter the
>> >> starting step)
>> >> 1028
>> >> 1029 ? d->createTransitionToPreviousStartingStep(d->StartingStep,
>> >> d->CurrentStep);
>> >> 1030
>> >> 1031 ? d->GoToStep = 0;
>> >> 1032 ? d->StartingStep->setStatusText("Attempt to go to the finish
>> >> step succeeded");
>> >> 1033 ? d->StartingStep = 0;
>> >> 1034
>> >> 1035 ? this->goFromGoToStepToStartingStep();
>> >> 1036 }
>> >> 1037
>> >>
>> >> Is this correct behavior and I understand the goToStep(x) call wrong
>> >> or is this a bug?
>> >>
>> >> Cheers,
>> >> Daniel
>> >> _______________________________________________
>> >> Ctk-developers mailing list
>> >> Ctk-developers at commontk.org
>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >
>> >
>> >
>> > --
>> > Danielle Pace, M.ESc.
>> > Research and Development Engineer
>> > Kitware Inc.,
>> > North Carolina Office
>> > www.kitware.com
>> > 919-969-6990 X 319
>> >
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
> www.kitware.com
> 919-969-6990 X 319
>


From haehn at bwh.harvard.edu  Fri Jun  3 19:57:23 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 15:57:23 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
	<BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>
	<BANLkTimJ58=syYEvZqNc7YNT7G0xNZBg_A@mail.gmail.com>
	<BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
Message-ID: <BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>

Hi Danielle,

thanks to your response I checked it again and you are right: it was
my mistake :)

If the name and description of the steps are set before the steps are
added to the workflow, it all works as expected.

I set the stuff after adding it and of course it did not recognize it
until the signal was fired.

Thanks for the great work!

Cheers,
Daniel

On Fri, Jun 3, 2011 at 3:44 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
> Hi Daniel,
> You shouldn't have to call the currenStepChanged signal directly. ?Can you
> please post your bit of code that is not working to the CTK developers
> mailing list, and I will take a look?
> Thanks,
>
> Danielle
>
> On Fri, Jun 3, 2011 at 3:21 PM, Julien Finet <julien.finet at kitware.com>
> wrote:
>>
>>
>> ---------- Forwarded message ----------
>> From: Daniel Haehn <haehn at bwh.harvard.edu>
>> Date: Fri, Jun 3, 2011 at 3:13 PM
>> Subject: Re: [Ctk-developers] Setting name and description of
>> ctkWorkflowWidgetStep
>> To: ctk-developers at commontk.org
>>
>>
>> Hi guys,
>>
>> I got it:
>>
>> the ctkWorkflowGroupBox listens for the currentStepChanged signal to
>> update the title (==name) and subTitle (==description) for the current
>> step.
>>
>> So maybe it makes sense to invoke this signal when the
>> ctkWorkflow.start() gets called?
>>
>> If I do it manually like that, it works:
>>
>> ? ? ?self.workflow.start()
>> ? ? ?self.workflow.currentStepChanged(steps[0])
>>
>> Cheers,
>> Daniel
>>
>> On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> wrote:
>> > Hi devels,
>> >
>> > in C++ I can do
>> >
>> > ?this->setName("Title of step");
>> > ?this->setDescription("Instructions");
>> >
>> > in a ctkWorkflowWidgetStep, which then get nicely displayed in the
>> > associated ctkWorkflowGroupBox.
>> >
>> > In Python, the same methods can be accessed but the
>> > ctkWorkflowGroupBox does not seem to show them.
>> >
>> > Do I need to trigger an update on the box or something similar?
>> >
>> > Thank you!
>> > Daniel
>> >
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
> www.kitware.com
> 919-969-6990 X 319
>


From danielle.pace at kitware.com  Fri Jun  3 20:13:38 2011
From: danielle.pace at kitware.com (Danielle Pace)
Date: Fri, 3 Jun 2011 16:13:38 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
	<BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>
	<BANLkTimJ58=syYEvZqNc7YNT7G0xNZBg_A@mail.gmail.com>
	<BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
	<BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>
Message-ID: <BANLkTinMLEtQvjRcOuq-f9VaV6NvCnteMw@mail.gmail.com>

Hi Daniel,

Actually, I think you exposed something important here.

I'm assuming, from python, that you wrote something similar to:

- create step 1 (with no name/description)
- create step 2 (with no name/description)
- create workflow widget and add steps
- display workflow widget
- set name/description for steps 1 and 2
- expected to see workflow update, and it didn't

Is this correct?  I think we are missing some signals on ctkWorkflowStep
such as NameChanged(QString) and DescriptionChanged(QString) that the
ctkWorkflowGroupBox should be listening to.

Perhaps you could still post your snippet of python code that wasn't
working, so I can verify that it is an ordering of commands that we should
support?

Thanks,

Danielle




On Fri, Jun 3, 2011 at 3:57 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi Danielle,
>
> thanks to your response I checked it again and you are right: it was
> my mistake :)
>
> If the name and description of the steps are set before the steps are
> added to the workflow, it all works as expected.
>
> I set the stuff after adding it and of course it did not recognize it
> until the signal was fired.
>
> Thanks for the great work!
>
> Cheers,
> Daniel
>
> On Fri, Jun 3, 2011 at 3:44 PM, Danielle Pace <danielle.pace at kitware.com>
> wrote:
> > Hi Daniel,
> > You shouldn't have to call the currenStepChanged signal directly.  Can
> you
> > please post your bit of code that is not working to the CTK developers
> > mailing list, and I will take a look?
> > Thanks,
> >
> > Danielle
> >
> > On Fri, Jun 3, 2011 at 3:21 PM, Julien Finet <julien.finet at kitware.com>
> > wrote:
> >>
> >>
> >> ---------- Forwarded message ----------
> >> From: Daniel Haehn <haehn at bwh.harvard.edu>
> >> Date: Fri, Jun 3, 2011 at 3:13 PM
> >> Subject: Re: [Ctk-developers] Setting name and description of
> >> ctkWorkflowWidgetStep
> >> To: ctk-developers at commontk.org
> >>
> >>
> >> Hi guys,
> >>
> >> I got it:
> >>
> >> the ctkWorkflowGroupBox listens for the currentStepChanged signal to
> >> update the title (==name) and subTitle (==description) for the current
> >> step.
> >>
> >> So maybe it makes sense to invoke this signal when the
> >> ctkWorkflow.start() gets called?
> >>
> >> If I do it manually like that, it works:
> >>
> >>      self.workflow.start()
> >>      self.workflow.currentStepChanged(steps[0])
> >>
> >> Cheers,
> >> Daniel
> >>
> >> On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn <haehn at bwh.harvard.edu>
> >> wrote:
> >> > Hi devels,
> >> >
> >> > in C++ I can do
> >> >
> >> >  this->setName("Title of step");
> >> >  this->setDescription("Instructions");
> >> >
> >> > in a ctkWorkflowWidgetStep, which then get nicely displayed in the
> >> > associated ctkWorkflowGroupBox.
> >> >
> >> > In Python, the same methods can be accessed but the
> >> > ctkWorkflowGroupBox does not seem to show them.
> >> >
> >> > Do I need to trigger an update on the box or something similar?
> >> >
> >> > Thank you!
> >> > Daniel
> >> >
> >> _______________________________________________
> >> Ctk-developers mailing list
> >> Ctk-developers at commontk.org
> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >>
> >
> >
> >
> > --
> > Danielle Pace, M.ESc.
> > Research and Development Engineer
> > Kitware Inc.,
> > North Carolina Office
> > www.kitware.com
> > 919-969-6990 X 319
> >
>



-- 
Danielle Pace, M.ESc.
Research and Development Engineer
Kitware Inc.,
North Carolina Office

www.kitware.com
919-969-6990 X 319
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/7a885189/attachment.htm>

From jchris.fillionr at kitware.com  Fri Jun  3 20:16:19 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Fri, 3 Jun 2011 16:16:19 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
	<BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
Message-ID: <BANLkTikqadOwLBmLU5Ktibfi3qx_1DHeaQ@mail.gmail.com>

Hi Daniel,

* A workflow is set of transitions.

* If you want to go from a Step X to a Step Y, a transition is required to
exists

* For conditional workflow, there is the concept of branch.

* Each time you click on finish, the pipeline (or workflow) is executed.
Upon success, it returns to the step from where the transition to the
"finish" step was done.  Having a different behavior would make sens. I
propose the following:
   * Add a property named "GoBackToOriginStepUponSuccess" to workflow.
(also open to suggestion regarding the name)
   * By default, this property would be False

Would be great if you could implement the described behavior. To do so:
   1) fork CTK on github
   2) Create a topic branch named
"add-GoBackToOriginStepUponSuccess-workflow-feature" of master
   3) Commit both feature and updated test
   4) Check that test run locally
   5) Send an email to the list asking for review with a pointer to your
topic
   6) Danielle or myself will review your topic and integrate it into master

Note: Make also sure the property is exposed in python. You could play with
the following example to make sure everything is working as expected. See
https://github.com/jcfr/SlicerPy/blob/master/slicer-visual-workflow-example.py

Thanks
Jc

On Fri, Jun 3, 2011 at 3:52 PM, Danielle Pace <danielle.pace at kitware.com>wrote:

> Hi Daniel,
>
> CanGoToStep checks to see if a path exists between two steps - so it is not
> helpful in your case.
>
> Right now, there is no method to go to a specific step without
> transitioning back.  We could add a property that dictates whether to go
> back to the starting step after successfully getting to a 'finish' step...
>
> Thanks,
>
> Danielle
>
>
>
>
>
> On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu>wrote:
>
>> Hi Danielle,
>>
>> hm this does not make totally sense to me. What about the canGoToStep
>> method - this only checks for next step?
>>
>> Wouldn't it make more sense to have the behavior I observed and you
>> described for the canGoToStep call and do not go back when you call
>> goToStep?
>>
>> Or is there another method I could use to go from step 1 to step 4, f.e.?
>>
>> Thanks
>> Daniel
>>
>> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace <danielle.pace at kitware.com>
>> wrote:
>> > Hi Daniel,
>> > This is the intended behavior when using 'goToStep'.
>> > The idea is that, partway through the workflow, you could execute the
>> steps
>> > until a 'finish step' was reached - using the default parameters for
>> those
>> > steps.  Then you can evaluate the results that occur with default
>> > parameters.  If you get to the 'finish' step successfully, it puts you
>> back
>> > where you started - so that you can continue to step through and alter
>> > parameters if you like.
>> > If you do not get to the 'finish' step successfully (i.e. default
>> parameters
>> > produce errors on your data or because of the parameters you set in
>> prior
>> > steps), it will stop at the error.
>> > Hope that helps,
>> >
>> > Danielle
>> >
>> >
>> >
>> >
>> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> wrote:
>> >>
>> >> Hi guys,
>> >>
>> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>> >> along the workflow (according to onEntry, validate and onExit calls).
>> >>
>> >> But then, after reaching step x, it goes directly back to where I am
>> >> coming from - again along the workflow. This goes so fast so it seems
>> >> there is no movement at all.
>> >>
>> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>> >> right? After reaching the 'finish step' (which in my understanding is
>> >> the target step), it goes back to the starting step.
>> >>
>> >>
>> >> 1019 //
>> >>
>> --------------------------------------------------------------------------
>> >> 1020 void ctkWorkflow::goToStepSucceeded()
>> >> 1021 {
>> >> 1022   Q_D(ctkWorkflow);
>> >> 1023
>> >> 1024   logger.debug("goToStepSucceeded");
>> >> 1025
>> >> 1026   // after success, go back to the step at which we begin looking
>> for
>> >> 1027   // the finish step (will exit the current step and enter the
>> >> starting step)
>> >> 1028
>> >> 1029   d->createTransitionToPreviousStartingStep(d->StartingStep,
>> >> d->CurrentStep);
>> >> 1030
>> >> 1031   d->GoToStep = 0;
>> >> 1032   d->StartingStep->setStatusText("Attempt to go to the finish
>> >> step succeeded");
>> >> 1033   d->StartingStep = 0;
>> >> 1034
>> >> 1035   this->goFromGoToStepToStartingStep();
>> >> 1036 }
>> >> 1037
>> >>
>> >> Is this correct behavior and I understand the goToStep(x) call wrong
>> >> or is this a bug?
>> >>
>> >> Cheers,
>> >> Daniel
>> >> _______________________________________________
>> >> Ctk-developers mailing list
>> >> Ctk-developers at commontk.org
>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >
>> >
>> >
>> > --
>> > Danielle Pace, M.ESc.
>> > Research and Development Engineer
>> > Kitware Inc.,
>> > North Carolina Office
>> > www.kitware.com
>> > 919-969-6990 X 319
>> >
>>
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
>
> www.kitware.com
> 919-969-6990 X 319
>
>
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>


-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/d1ce92c4/attachment.htm>

From danielle.pace at kitware.com  Fri Jun  3 20:22:53 2011
From: danielle.pace at kitware.com (Danielle Pace)
Date: Fri, 3 Jun 2011 16:22:53 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTikqadOwLBmLU5Ktibfi3qx_1DHeaQ@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
	<BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
	<BANLkTikqadOwLBmLU5Ktibfi3qx_1DHeaQ@mail.gmail.com>
Message-ID: <BANLkTimzqNjgS73x91_Raf2xCvheXcy8dg@mail.gmail.com>

FYI: The function ctkWorkflow::goToStepSucceeded() is called when
ctkWorkflow::goToStep()... succeeds =)

If my memory serves me correctly, you should have to simply make the two
lines:

  d->createTransitionToPreviousStartingStep(d->StartingStep, d->CurrentStep);

   this->goFromGoToStepToStartingStep();


conditional on the value of the new property, and the housekeeping:

  d->GoToStep = 0;

  d->StartingStep->setStatusText("Attempt to go to the finish step succeeded");

  d->StartingStep = 0;


should be done regardless of the property's value.

Thanks,

Danielle


On Fri, Jun 3, 2011 at 4:16 PM, Jean-Christophe Fillion-Robin <
jchris.fillionr at kitware.com> wrote:

> Hi Daniel,
>
> * A workflow is set of transitions.
>
> * If you want to go from a Step X to a Step Y, a transition is required to
> exists
>
> * For conditional workflow, there is the concept of branch.
>
> * Each time you click on finish, the pipeline (or workflow) is executed.
> Upon success, it returns to the step from where the transition to the
> "finish" step was done.  Having a different behavior would make sens. I
> propose the following:
>    * Add a property named "GoBackToOriginStepUponSuccess" to workflow.
> (also open to suggestion regarding the name)
>    * By default, this property would be False
>
> Would be great if you could implement the described behavior. To do so:
>    1) fork CTK on github
>    2) Create a topic branch named
> "add-GoBackToOriginStepUponSuccess-workflow-feature" of master
>    3) Commit both feature and updated test
>    4) Check that test run locally
>    5) Send an email to the list asking for review with a pointer to your
> topic
>    6) Danielle or myself will review your topic and integrate it into
> master
>
> Note: Make also sure the property is exposed in python. You could play with
> the following example to make sure everything is working as expected. See
> https://github.com/jcfr/SlicerPy/blob/master/slicer-visual-workflow-example.py
>
> Thanks
> Jc
>
>
> On Fri, Jun 3, 2011 at 3:52 PM, Danielle Pace <danielle.pace at kitware.com>wrote:
>
>> Hi Daniel,
>>
>> CanGoToStep checks to see if a path exists between two steps - so it is
>> not helpful in your case.
>>
>> Right now, there is no method to go to a specific step without
>> transitioning back.  We could add a property that dictates whether to go
>> back to the starting step after successfully getting to a 'finish' step...
>>
>> Thanks,
>>
>> Danielle
>>
>>
>>
>>
>>
>> On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu>wrote:
>>
>>> Hi Danielle,
>>>
>>> hm this does not make totally sense to me. What about the canGoToStep
>>> method - this only checks for next step?
>>>
>>> Wouldn't it make more sense to have the behavior I observed and you
>>> described for the canGoToStep call and do not go back when you call
>>> goToStep?
>>>
>>> Or is there another method I could use to go from step 1 to step 4, f.e.?
>>>
>>> Thanks
>>> Daniel
>>>
>>> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace <danielle.pace at kitware.com>
>>> wrote:
>>> > Hi Daniel,
>>> > This is the intended behavior when using 'goToStep'.
>>> > The idea is that, partway through the workflow, you could execute the
>>> steps
>>> > until a 'finish step' was reached - using the default parameters for
>>> those
>>> > steps.  Then you can evaluate the results that occur with default
>>> > parameters.  If you get to the 'finish' step successfully, it puts you
>>> back
>>> > where you started - so that you can continue to step through and alter
>>> > parameters if you like.
>>> > If you do not get to the 'finish' step successfully (i.e. default
>>> parameters
>>> > produce errors on your data or because of the parameters you set in
>>> prior
>>> > steps), it will stop at the error.
>>> > Hope that helps,
>>> >
>>> > Danielle
>>> >
>>> >
>>> >
>>> >
>>> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>>> wrote:
>>> >>
>>> >> Hi guys,
>>> >>
>>> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>>> >> along the workflow (according to onEntry, validate and onExit calls).
>>> >>
>>> >> But then, after reaching step x, it goes directly back to where I am
>>> >> coming from - again along the workflow. This goes so fast so it seems
>>> >> there is no movement at all.
>>> >>
>>> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>>> >> right? After reaching the 'finish step' (which in my understanding is
>>> >> the target step), it goes back to the starting step.
>>> >>
>>> >>
>>> >> 1019 //
>>> >>
>>> --------------------------------------------------------------------------
>>> >> 1020 void ctkWorkflow::goToStepSucceeded()
>>> >> 1021 {
>>> >> 1022   Q_D(ctkWorkflow);
>>> >> 1023
>>> >> 1024   logger.debug("goToStepSucceeded");
>>> >> 1025
>>> >> 1026   // after success, go back to the step at which we begin looking
>>> for
>>> >> 1027   // the finish step (will exit the current step and enter the
>>> >> starting step)
>>> >> 1028
>>> >> 1029   d->createTransitionToPreviousStartingStep(d->StartingStep,
>>> >> d->CurrentStep);
>>> >> 1030
>>> >> 1031   d->GoToStep = 0;
>>> >> 1032   d->StartingStep->setStatusText("Attempt to go to the finish
>>> >> step succeeded");
>>> >> 1033   d->StartingStep = 0;
>>> >> 1034
>>> >> 1035   this->goFromGoToStepToStartingStep();
>>> >> 1036 }
>>> >> 1037
>>> >>
>>> >> Is this correct behavior and I understand the goToStep(x) call wrong
>>> >> or is this a bug?
>>> >>
>>> >> Cheers,
>>> >> Daniel
>>> >> _______________________________________________
>>> >> Ctk-developers mailing list
>>> >> Ctk-developers at commontk.org
>>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>> >
>>> >
>>> >
>>> > --
>>> > Danielle Pace, M.ESc.
>>> > Research and Development Engineer
>>> > Kitware Inc.,
>>> > North Carolina Office
>>> > www.kitware.com
>>> > 919-969-6990 X 319
>>> >
>>>
>>
>>
>>
>> --
>> Danielle Pace, M.ESc.
>> Research and Development Engineer
>> Kitware Inc.,
>> North Carolina Office
>>
>> www.kitware.com
>> 919-969-6990 X 319
>>
>>
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
>
>
> --
> +1 919 869 8849
>
>


-- 
Danielle Pace, M.ESc.
Research and Development Engineer
Kitware Inc.,
North Carolina Office

www.kitware.com
919-969-6990 X 319
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110603/25ccd6be/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun  3 20:31:25 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 3 Jun 2011 16:31:25 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTimzqNjgS73x91_Raf2xCvheXcy8dg@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
	<BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
	<BANLkTikqadOwLBmLU5Ktibfi3qx_1DHeaQ@mail.gmail.com>
	<BANLkTimzqNjgS73x91_Raf2xCvheXcy8dg@mail.gmail.com>
Message-ID: <BANLkTi=9sXbLDJc1D6j0U6tscXdZnUEq2w@mail.gmail.com>

Hi guys,

this makes sense to me. I will give it a shot at the weekend and let you know.

Cheers,
Daniel

On Fri, Jun 3, 2011 at 4:22 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
> FYI: The function ctkWorkflow::goToStepSucceeded() is called when
> ctkWorkflow::goToStep()... succeeds =)
> If my memory serves me correctly, you should have to simply make the two
> lines:
>
>   d->createTransitionToPreviousStartingStep(d->StartingStep,
> d->CurrentStep);
>
>   this->goFromGoToStepToStartingStep();
>
> conditional on the value of the new property, and the housekeeping:
>
>   d->GoToStep = 0;
>
>   d->StartingStep->setStatusText("Attempt to go to the finish step
> succeeded");
>
>   d->StartingStep = 0;
>
> should be done regardless of the property's value.
> Thanks,
>
> Danielle
>
> On Fri, Jun 3, 2011 at 4:16 PM, Jean-Christophe Fillion-Robin
> <jchris.fillionr at kitware.com> wrote:
>>
>> Hi Daniel,
>>
>> * A workflow is set of transitions.
>>
>> * If you want to go from a Step X to a Step Y, a transition is required to
>> exists
>>
>> * For conditional workflow, there is the concept of branch.
>>
>> * Each time you click on finish, the pipeline (or workflow) is executed.
>> Upon success, it returns to the step from where the transition to the
>> "finish" step was done.? Having a different behavior would make sens. I
>> propose the following:
>> ?? * Add a property named "GoBackToOriginStepUponSuccess" to workflow.
>> (also open to suggestion regarding the name)
>> ?? * By default, this property would be False
>>
>> Would be great if you could implement the described behavior. To do so:
>> ?? 1) fork CTK on github
>> ?? 2) Create a topic branch named
>> "add-GoBackToOriginStepUponSuccess-workflow-feature" of master
>> ?? 3) Commit both feature and updated test
>> ?? 4) Check that test run locally
>> ?? 5) Send an email to the list asking for review with a pointer to your
>> topic
>> ?? 6) Danielle or myself will review your topic and integrate it into
>> master
>>
>> Note: Make also sure the property is exposed in python. You could play
>> with the following example to make sure everything is working as expected.
>> See
>> https://github.com/jcfr/SlicerPy/blob/master/slicer-visual-workflow-example.py
>>
>> Thanks
>> Jc
>>
>> On Fri, Jun 3, 2011 at 3:52 PM, Danielle Pace <danielle.pace at kitware.com>
>> wrote:
>>>
>>> Hi Daniel,
>>> CanGoToStep checks to see if a path exists between two steps - so it is
>>> not helpful in your case.
>>> Right now, there is no method to go to a specific step without
>>> transitioning back. ?We could add a property that dictates whether to go
>>> back to the starting step after successfully getting to a 'finish' step...
>>> Thanks,
>>>
>>> Danielle
>>>
>>>
>>>
>>>
>>> On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>>> wrote:
>>>>
>>>> Hi Danielle,
>>>>
>>>> hm this does not make totally sense to me. What about the canGoToStep
>>>> method - this only checks for next step?
>>>>
>>>> Wouldn't it make more sense to have the behavior I observed and you
>>>> described for the canGoToStep call and do not go back when you call
>>>> goToStep?
>>>>
>>>> Or is there another method I could use to go from step 1 to step 4,
>>>> f.e.?
>>>>
>>>> Thanks
>>>> Daniel
>>>>
>>>> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace
>>>> <danielle.pace at kitware.com> wrote:
>>>> > Hi Daniel,
>>>> > This is the intended behavior when using 'goToStep'.
>>>> > The idea is that, partway through the workflow, you could execute the
>>>> > steps
>>>> > until a 'finish step' was reached - using the default parameters for
>>>> > those
>>>> > steps. ?Then you can evaluate the results that occur with default
>>>> > parameters. ?If you get to the 'finish' step successfully, it puts you
>>>> > back
>>>> > where you started - so that you can continue to step through and alter
>>>> > parameters if you like.
>>>> > If you do not get to the 'finish' step successfully (i.e. default
>>>> > parameters
>>>> > produce errors on your data or because of the parameters you set in
>>>> > prior
>>>> > steps), it will stop at the error.
>>>> > Hope that helps,
>>>> >
>>>> > Danielle
>>>> >
>>>> >
>>>> >
>>>> >
>>>> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>>>> > wrote:
>>>> >>
>>>> >> Hi guys,
>>>> >>
>>>> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>>>> >> along the workflow (according to onEntry, validate and onExit calls).
>>>> >>
>>>> >> But then, after reaching step x, it goes directly back to where I am
>>>> >> coming from - again along the workflow. This goes so fast so it seems
>>>> >> there is no movement at all.
>>>> >>
>>>> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>>>> >> right? After reaching the 'finish step' (which in my understanding is
>>>> >> the target step), it goes back to the starting step.
>>>> >>
>>>> >>
>>>> >> 1019 //
>>>> >>
>>>> >> --------------------------------------------------------------------------
>>>> >> 1020 void ctkWorkflow::goToStepSucceeded()
>>>> >> 1021 {
>>>> >> 1022 ? Q_D(ctkWorkflow);
>>>> >> 1023
>>>> >> 1024 ? logger.debug("goToStepSucceeded");
>>>> >> 1025
>>>> >> 1026 ? // after success, go back to the step at which we begin
>>>> >> looking for
>>>> >> 1027 ? // the finish step (will exit the current step and enter the
>>>> >> starting step)
>>>> >> 1028
>>>> >> 1029 ? d->createTransitionToPreviousStartingStep(d->StartingStep,
>>>> >> d->CurrentStep);
>>>> >> 1030
>>>> >> 1031 ? d->GoToStep = 0;
>>>> >> 1032 ? d->StartingStep->setStatusText("Attempt to go to the finish
>>>> >> step succeeded");
>>>> >> 1033 ? d->StartingStep = 0;
>>>> >> 1034
>>>> >> 1035 ? this->goFromGoToStepToStartingStep();
>>>> >> 1036 }
>>>> >> 1037
>>>> >>
>>>> >> Is this correct behavior and I understand the goToStep(x) call wrong
>>>> >> or is this a bug?
>>>> >>
>>>> >> Cheers,
>>>> >> Daniel
>>>> >> _______________________________________________
>>>> >> Ctk-developers mailing list
>>>> >> Ctk-developers at commontk.org
>>>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>> >
>>>> >
>>>> >
>>>> > --
>>>> > Danielle Pace, M.ESc.
>>>> > Research and Development Engineer
>>>> > Kitware Inc.,
>>>> > North Carolina Office
>>>> > www.kitware.com
>>>> > 919-969-6990 X 319
>>>> >
>>>
>>>
>>>
>>> --
>>> Danielle Pace, M.ESc.
>>> Research and Development Engineer
>>> Kitware Inc.,
>>> North Carolina Office
>>> www.kitware.com
>>> 919-969-6990 X 319
>>>
>>> _______________________________________________
>>> Ctk-developers mailing list
>>> Ctk-developers at commontk.org
>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>
>>
>>
>>
>> --
>> +1 919 869 8849
>>
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
> www.kitware.com
> 919-969-6990 X 319
>


From haehn at bwh.harvard.edu  Sun Jun  5 02:32:39 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Sat, 4 Jun 2011 22:32:39 -0400
Subject: [Ctk-developers] Another ctkWorkflow thingie: goToStep(..) goes
 always back to starting step
In-Reply-To: <BANLkTi=9sXbLDJc1D6j0U6tscXdZnUEq2w@mail.gmail.com>
References: <BANLkTi=QZsPEYaPX=cM-3upt_VP4M5rWRA@mail.gmail.com>
	<BANLkTi=eYgHW3apnyNh53Vb0cMHZXLa2LA@mail.gmail.com>
	<BANLkTinre-2=jjxtYOQ1m1SCmmZQpFyytQ@mail.gmail.com>
	<BANLkTi=+N37C8DK6MmW-29UwE3Qky9kxUA@mail.gmail.com>
	<BANLkTikqadOwLBmLU5Ktibfi3qx_1DHeaQ@mail.gmail.com>
	<BANLkTimzqNjgS73x91_Raf2xCvheXcy8dg@mail.gmail.com>
	<BANLkTi=9sXbLDJc1D6j0U6tscXdZnUEq2w@mail.gmail.com>
Message-ID: <BANLkTikGq1uuJndTPMuwzBXF=5VgMAAxRw@mail.gmail.com>

Hi guys,

please review the topic in my github CTK fork:

https://github.com/haehn/CTK/tree/add-GoBackToOriginStepUponSuccess-workflow-feature

Thanks,
Daniel

On Fri, Jun 3, 2011 at 4:31 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
> Hi guys,
>
> this makes sense to me. I will give it a shot at the weekend and let you know.
>
> Cheers,
> Daniel
>
> On Fri, Jun 3, 2011 at 4:22 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
>> FYI: The function ctkWorkflow::goToStepSucceeded() is called when
>> ctkWorkflow::goToStep()... succeeds =)
>> If my memory serves me correctly, you should have to simply make the two
>> lines:
>>
>> ? d->createTransitionToPreviousStartingStep(d->StartingStep,
>> d->CurrentStep);
>>
>> ? this->goFromGoToStepToStartingStep();
>>
>> conditional on the value of the new property, and the housekeeping:
>>
>> ? d->GoToStep = 0;
>>
>> ? d->StartingStep->setStatusText("Attempt to go to the finish step
>> succeeded");
>>
>> ? d->StartingStep = 0;
>>
>> should be done regardless of the property's value.
>> Thanks,
>>
>> Danielle
>>
>> On Fri, Jun 3, 2011 at 4:16 PM, Jean-Christophe Fillion-Robin
>> <jchris.fillionr at kitware.com> wrote:
>>>
>>> Hi Daniel,
>>>
>>> * A workflow is set of transitions.
>>>
>>> * If you want to go from a Step X to a Step Y, a transition is required to
>>> exists
>>>
>>> * For conditional workflow, there is the concept of branch.
>>>
>>> * Each time you click on finish, the pipeline (or workflow) is executed.
>>> Upon success, it returns to the step from where the transition to the
>>> "finish" step was done.? Having a different behavior would make sens. I
>>> propose the following:
>>> ?? * Add a property named "GoBackToOriginStepUponSuccess" to workflow.
>>> (also open to suggestion regarding the name)
>>> ?? * By default, this property would be False
>>>
>>> Would be great if you could implement the described behavior. To do so:
>>> ?? 1) fork CTK on github
>>> ?? 2) Create a topic branch named
>>> "add-GoBackToOriginStepUponSuccess-workflow-feature" of master
>>> ?? 3) Commit both feature and updated test
>>> ?? 4) Check that test run locally
>>> ?? 5) Send an email to the list asking for review with a pointer to your
>>> topic
>>> ?? 6) Danielle or myself will review your topic and integrate it into
>>> master
>>>
>>> Note: Make also sure the property is exposed in python. You could play
>>> with the following example to make sure everything is working as expected.
>>> See
>>> https://github.com/jcfr/SlicerPy/blob/master/slicer-visual-workflow-example.py
>>>
>>> Thanks
>>> Jc
>>>
>>> On Fri, Jun 3, 2011 at 3:52 PM, Danielle Pace <danielle.pace at kitware.com>
>>> wrote:
>>>>
>>>> Hi Daniel,
>>>> CanGoToStep checks to see if a path exists between two steps - so it is
>>>> not helpful in your case.
>>>> Right now, there is no method to go to a specific step without
>>>> transitioning back. ?We could add a property that dictates whether to go
>>>> back to the starting step after successfully getting to a 'finish' step...
>>>> Thanks,
>>>>
>>>> Danielle
>>>>
>>>>
>>>>
>>>>
>>>> On Fri, Jun 3, 2011 at 3:45 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>>>> wrote:
>>>>>
>>>>> Hi Danielle,
>>>>>
>>>>> hm this does not make totally sense to me. What about the canGoToStep
>>>>> method - this only checks for next step?
>>>>>
>>>>> Wouldn't it make more sense to have the behavior I observed and you
>>>>> described for the canGoToStep call and do not go back when you call
>>>>> goToStep?
>>>>>
>>>>> Or is there another method I could use to go from step 1 to step 4,
>>>>> f.e.?
>>>>>
>>>>> Thanks
>>>>> Daniel
>>>>>
>>>>> On Fri, Jun 3, 2011 at 3:40 PM, Danielle Pace
>>>>> <danielle.pace at kitware.com> wrote:
>>>>> > Hi Daniel,
>>>>> > This is the intended behavior when using 'goToStep'.
>>>>> > The idea is that, partway through the workflow, you could execute the
>>>>> > steps
>>>>> > until a 'finish step' was reached - using the default parameters for
>>>>> > those
>>>>> > steps. ?Then you can evaluate the results that occur with default
>>>>> > parameters. ?If you get to the 'finish' step successfully, it puts you
>>>>> > back
>>>>> > where you started - so that you can continue to step through and alter
>>>>> > parameters if you like.
>>>>> > If you do not get to the 'finish' step successfully (i.e. default
>>>>> > parameters
>>>>> > produce errors on your data or because of the parameters you set in
>>>>> > prior
>>>>> > steps), it will stop at the error.
>>>>> > Hope that helps,
>>>>> >
>>>>> > Danielle
>>>>> >
>>>>> >
>>>>> >
>>>>> >
>>>>> > On Fri, Jun 3, 2011 at 3:34 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>>>>> > wrote:
>>>>> >>
>>>>> >> Hi guys,
>>>>> >>
>>>>> >> if I use ctkWorkflow.goToStep(x), the workflow goes to the step x
>>>>> >> along the workflow (according to onEntry, validate and onExit calls).
>>>>> >>
>>>>> >> But then, after reaching step x, it goes directly back to where I am
>>>>> >> coming from - again along the workflow. This goes so fast so it seems
>>>>> >> there is no movement at all.
>>>>> >>
>>>>> >> I digged a little and it seems line 1035 in ctkWorkflow.cpp is not
>>>>> >> right? After reaching the 'finish step' (which in my understanding is
>>>>> >> the target step), it goes back to the starting step.
>>>>> >>
>>>>> >>
>>>>> >> 1019 //
>>>>> >>
>>>>> >> --------------------------------------------------------------------------
>>>>> >> 1020 void ctkWorkflow::goToStepSucceeded()
>>>>> >> 1021 {
>>>>> >> 1022 ? Q_D(ctkWorkflow);
>>>>> >> 1023
>>>>> >> 1024 ? logger.debug("goToStepSucceeded");
>>>>> >> 1025
>>>>> >> 1026 ? // after success, go back to the step at which we begin
>>>>> >> looking for
>>>>> >> 1027 ? // the finish step (will exit the current step and enter the
>>>>> >> starting step)
>>>>> >> 1028
>>>>> >> 1029 ? d->createTransitionToPreviousStartingStep(d->StartingStep,
>>>>> >> d->CurrentStep);
>>>>> >> 1030
>>>>> >> 1031 ? d->GoToStep = 0;
>>>>> >> 1032 ? d->StartingStep->setStatusText("Attempt to go to the finish
>>>>> >> step succeeded");
>>>>> >> 1033 ? d->StartingStep = 0;
>>>>> >> 1034
>>>>> >> 1035 ? this->goFromGoToStepToStartingStep();
>>>>> >> 1036 }
>>>>> >> 1037
>>>>> >>
>>>>> >> Is this correct behavior and I understand the goToStep(x) call wrong
>>>>> >> or is this a bug?
>>>>> >>
>>>>> >> Cheers,
>>>>> >> Daniel
>>>>> >> _______________________________________________
>>>>> >> Ctk-developers mailing list
>>>>> >> Ctk-developers at commontk.org
>>>>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>>> >
>>>>> >
>>>>> >
>>>>> > --
>>>>> > Danielle Pace, M.ESc.
>>>>> > Research and Development Engineer
>>>>> > Kitware Inc.,
>>>>> > North Carolina Office
>>>>> > www.kitware.com
>>>>> > 919-969-6990 X 319
>>>>> >
>>>>
>>>>
>>>>
>>>> --
>>>> Danielle Pace, M.ESc.
>>>> Research and Development Engineer
>>>> Kitware Inc.,
>>>> North Carolina Office
>>>> www.kitware.com
>>>> 919-969-6990 X 319
>>>>
>>>> _______________________________________________
>>>> Ctk-developers mailing list
>>>> Ctk-developers at commontk.org
>>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>>
>>>
>>>
>>>
>>> --
>>> +1 919 869 8849
>>>
>>
>>
>>
>> --
>> Danielle Pace, M.ESc.
>> Research and Development Engineer
>> Kitware Inc.,
>> North Carolina Office
>> www.kitware.com
>> 919-969-6990 X 319
>>
>


From haehn at bwh.harvard.edu  Sun Jun  5 02:41:20 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Sat, 4 Jun 2011 22:41:20 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTinMLEtQvjRcOuq-f9VaV6NvCnteMw@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
	<BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>
	<BANLkTimJ58=syYEvZqNc7YNT7G0xNZBg_A@mail.gmail.com>
	<BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
	<BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>
	<BANLkTinMLEtQvjRcOuq-f9VaV6NvCnteMw@mail.gmail.com>
Message-ID: <BANLkTi=sYr6megrCtJR3b2nE2Tj4mpsdTw@mail.gmail.com>

Hi Danielle,

that is absolutely correct!

Maybe it would be easiest to just fire a
d->workflow->currentStepChanged(this) signal when
ctkWorkflowStep::setName or setDescription are called? Then, the
groupbox would be automatically updated.

Nevertheless, it has no high priority for me since I now set the name
and description of the step in its constructor - before adding it to
the workflow.

Cheers,
Daniel

On Fri, Jun 3, 2011 at 4:13 PM, Danielle Pace <danielle.pace at kitware.com> wrote:
> Hi Daniel,
> Actually, I think you exposed something important here.
> I'm assuming, from python, that you wrote something similar to:
> - create step 1 (with no name/description)
> - create step 2 (with no name/description)
> - create workflow widget and add steps
> - display workflow widget
> - set name/description for steps 1 and 2
> - expected to see workflow update, and it didn't
> Is this correct? ?I think we are missing some signals on ctkWorkflowStep
> such as NameChanged(QString) and DescriptionChanged(QString) that the
> ctkWorkflowGroupBox should be listening to.
> Perhaps you could still post your snippet of python code that wasn't
> working, so I can verify that it is an ordering of commands that we should
> support?
> Thanks,
>
> Danielle
>
>
>
> On Fri, Jun 3, 2011 at 3:57 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
>>
>> Hi Danielle,
>>
>> thanks to your response I checked it again and you are right: it was
>> my mistake :)
>>
>> If the name and description of the steps are set before the steps are
>> added to the workflow, it all works as expected.
>>
>> I set the stuff after adding it and of course it did not recognize it
>> until the signal was fired.
>>
>> Thanks for the great work!
>>
>> Cheers,
>> Daniel
>>
>> On Fri, Jun 3, 2011 at 3:44 PM, Danielle Pace <danielle.pace at kitware.com>
>> wrote:
>> > Hi Daniel,
>> > You shouldn't have to call the currenStepChanged signal directly. ?Can
>> > you
>> > please post your bit of code that is not working to the CTK developers
>> > mailing list, and I will take a look?
>> > Thanks,
>> >
>> > Danielle
>> >
>> > On Fri, Jun 3, 2011 at 3:21 PM, Julien Finet <julien.finet at kitware.com>
>> > wrote:
>> >>
>> >>
>> >> ---------- Forwarded message ----------
>> >> From: Daniel Haehn <haehn at bwh.harvard.edu>
>> >> Date: Fri, Jun 3, 2011 at 3:13 PM
>> >> Subject: Re: [Ctk-developers] Setting name and description of
>> >> ctkWorkflowWidgetStep
>> >> To: ctk-developers at commontk.org
>> >>
>> >>
>> >> Hi guys,
>> >>
>> >> I got it:
>> >>
>> >> the ctkWorkflowGroupBox listens for the currentStepChanged signal to
>> >> update the title (==name) and subTitle (==description) for the current
>> >> step.
>> >>
>> >> So maybe it makes sense to invoke this signal when the
>> >> ctkWorkflow.start() gets called?
>> >>
>> >> If I do it manually like that, it works:
>> >>
>> >> ? ? ?self.workflow.start()
>> >> ? ? ?self.workflow.currentStepChanged(steps[0])
>> >>
>> >> Cheers,
>> >> Daniel
>> >>
>> >> On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> >> wrote:
>> >> > Hi devels,
>> >> >
>> >> > in C++ I can do
>> >> >
>> >> > ?this->setName("Title of step");
>> >> > ?this->setDescription("Instructions");
>> >> >
>> >> > in a ctkWorkflowWidgetStep, which then get nicely displayed in the
>> >> > associated ctkWorkflowGroupBox.
>> >> >
>> >> > In Python, the same methods can be accessed but the
>> >> > ctkWorkflowGroupBox does not seem to show them.
>> >> >
>> >> > Do I need to trigger an update on the box or something similar?
>> >> >
>> >> > Thank you!
>> >> > Daniel
>> >> >
>> >> _______________________________________________
>> >> Ctk-developers mailing list
>> >> Ctk-developers at commontk.org
>> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >>
>> >
>> >
>> >
>> > --
>> > Danielle Pace, M.ESc.
>> > Research and Development Engineer
>> > Kitware Inc.,
>> > North Carolina Office
>> > www.kitware.com
>> > 919-969-6990 X 319
>> >
>
>
>
> --
> Danielle Pace, M.ESc.
> Research and Development Engineer
> Kitware Inc.,
> North Carolina Office
> www.kitware.com
> 919-969-6990 X 319
>


From haehn at bwh.harvard.edu  Mon Jun  6 15:06:01 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Mon, 6 Jun 2011 11:06:01 -0400
Subject: [Ctk-developers] ctkWorkflowButtonBoxWidget does not show text for
	'next' button
Message-ID: <BANLkTimzfrffr2LX7WG00qkSSD-WAHq2Bw@mail.gmail.com>

Hi devels,

I noticed that the text of the next button is not shown in the
ctkWorkflowButtonBoxWidget on Snow Leopard.

The underlying ctk code uses
this->NextButton->setLayoutDirection(Qt::RightToLeft);  to place the
icon on the right side of the text.

I think that this might be a misuse of the layout direction and causes
the problems. The back button is set up the same way as the next
button without the layout direction and works fine.

What do you guys think?

Thanks,
Daniel


From michael.caylus at siemens.com  Mon Jun  6 19:35:24 2011
From: michael.caylus at siemens.com (Caylus, Michael (SCR US))
Date: Mon, 6 Jun 2011 15:35:24 -0400
Subject: [Ctk-developers] data exchange in the DICOM Hosting implementation
 / QtSoap Limitation
Message-ID: <63D07F9AA364DB4C901F02EF93465C8D780878D789@USMLVA0E934MSX.ww017.siemens.net>

Hi Sascha,

First thanks a lot for the quick reply I got on my previous post.  I decided to go for the first approach and created my own CTK plugin as we should stay in the context of a hosting application. I managed to start and run my hosted application through the DICOM Application Host interface with this approach.

I am now looking into passing some dataset to the hosted application.  I wanted to start with something simple where the end-user select a folder that contains a list of DICOM files (one series) and pass it to my application. After choosing the folder; I am updating the AvailableData structure and then publish it to the hosted application when I am starting it. By the time the SOAP request is done I could see that the SOAP message will contains my list of uuids I created for one series but when the message is received (ctkSoapMessageProcessorList::process method) on the hosted application; the message is not the same and only contains one uuid now for the series. Is there any limitation currently within QtSoap to support arrays that you are aware of?  I have seen some discussion regarding on the ctk website (http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment) but I am not sure what the status here.

If there is still some restriction with QtSoap regarding the array; should I send multiple notifyDataAvailable messages to the application; one for each file I am willing to get on the application side. Is there any other best alternative beside my first approach?

Best,
Michael


-------------------------------------------------------------------------------------------------------------------------------
From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de>
Subject: Re: [Ctk-developers] How to wrap a Qt App into an hosted app
        confirming to DICOM Supplement 118 / Working Group 23
To: ctk-developers at commontk.org
Message-ID: <4DCD447F.7060402 at dkfz-heidelberg.de>
Content-Type: text/plain; charset="windows-1252"; Format="flowed"

Hi Michael,

Welcome to the CTK mailing list.

DICOM Application Hosting support in CTK is still work in progress, so please be aware that not everything will work yet. Especially the data exchange interface needs more work. Nevertheless, I hope my comments below will get you started to experiment with your own Qt based DICOM App. Comments, bug reports, patches etc. are also always welcome!

Best,
Sascha

On 05/13/2011 03:56 AM, Caylus, Michael (SCR US) wrote:
> Hello,
> I am new to the CTK SDK and would like to wrap my Qt application into
> an hosted app. I have seen such an example by turning on in the CMake
> configuration:  ctkExampleHost and ctkExampleHosted App.  I came up
> with the following questions after playing with it:
>
>    1. In order to make this wrap-up for my app; I was considering to
>       edit the org_commontk_dah_exampleapp plug-in which relies on the
>       org_commontk_dah_app and org_commontk_dah_core plug-in and keep
>       the ctkExampleHostedApp. Is this approach the right way to
>       proceed? This implies splitting my app into the hostedApp and a
>       CTK plugin; with most of the work to be done on the plug-in
>       level I guess. I have seen some doc explaining how to embed the
>       CTK plugin framework in my app too. Can I achieve the same
>       result with this approach and how?
>
As you have already seen, it is possible to use the plugin framework from inside your own application and this is also the preferred way (instead of modifying the existing CTK plug-ins directly). In the case of a DICOM App, I can think of two different approaches, depending on your requirements:

1.) Your app is a "pure" DICOM App, i.e. you do not need run it outside the context of a "Hosting Application"

This is the easier approach. You would create your own plug-in, equivalent to org.commontk.dah.exampleapp, which contains all your app code (logic, widgets, etc.). Additionally, you create your own ctkExampleHostedApp executable, which is essentially the same code as in the CTK repository (just adapt paths and plug-in names).

The
http://www.commontk.org/index.php/Documentation/CTK_Plugin_Framework:_Setting_up_a_project?action=purge
tutorial explains how to create a CTK-based project.

2.) You want to use your app both "stand-alone" and as a "hosted app"

This involves a little bit more knowledge about the plugin framework, but should also be easy to achieve. What the org.commontk.dah.exampleapp plug-in actually does, is to register a class containing the DICOM "application logic" as a service within the plugin-framework. You could either create this class (similar as the ctkExampleDicomAppLocig class) inside your current project (not in a separate plug-in) and register an instance by using the plugin framework context, or create a small plug-in containing this class and taking care of the registration itself. Then start the plugin framework from inside your executable, if you recognize the host and app url command line parameters (otherwise, create a QMainWindow yourself).

Depending on how fancy you would like to get concerning the minimization of dependencies of your app in stand-alone mode to the CTK libraries, you could think of a more complex set-up containing two small wrapper executables, one for "app mode" starting the CTK plugin framework and one for "stand-alone mode", both linked to a library containing your application gui and logic...

>    1. I couldn?t find to which QMainWindow is the button attached to
>       in the ctkExampleDicomAppLogic::do_something()function.It
>       doesn?t look like thectkExampleDicomAppLogic is heritating from
>       a widgetalso. Which object is responsible of the creation of the
>       mainwindow?
>

The org.commontk.dah.exampleapp plugin does not create a QMainWindow. It just create a button an shows it (moved and resized to respect the prescribed screen area from the host). In Qt, widgets become top-level windows if they don't have a parent widget and are "shown".

>    1. Is there any doc that explains how to create a CTK plug-in and
>       App from scratch. It would be probably better in my case to
>       create a new plugin instead of editing an existing one.
>

There is no doc about creating a plug-in yet, but you can easily create one by using the ctkPluginGenerator executable (you may have to enable it in CMake). It might have some rough edges though...

A CTK "app" is actually nothing special. Any executable can do, which links to the CTK libraries. The tutorial linked above explains how to use the plugin framework in an arbitrary executable.


>    1. How can I add additional includes directories through a
>       CMakeLists in the CTK framework.  I couldn?t find a way to
>       define additional directories in the CMakeList of a plug-in (no
>       include_directories). Is that define somewhere else?
>

Inside CTK itself, include directories are set-up auto-magically through our custom CMake dependency system. If you need to use external libraries in your own CTK plug-in (contained in your own project), just use standard CMake INCLUDE_DIRECTORIES (before calling the ctkMacroBuildPlugin macro) and LINK_LIBRARIES(${PROJECT_NAME} ...) (after the ctkMacroBuildPlugin call) commands.

Happy coding,

Sascha
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110513/539b8908/attachment-0001.htm>

------------------------------

_______________________________________________
Ctk-developers mailing list
Ctk-developers at commontk.org<mailto:Ctk-developers at commontk.org>
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers


End of Ctk-developers Digest, Vol 22, Issue 6
*********************************************


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110606/609acd9d/attachment.htm>

From julien.finet at kitware.com  Tue Jun  7 02:30:18 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Mon, 6 Jun 2011 22:30:18 -0400
Subject: [Ctk-developers] ctkWorkflowButtonBoxWidget does not show text
 for 'next' button
In-Reply-To: <BANLkTimzfrffr2LX7WG00qkSSD-WAHq2Bw@mail.gmail.com>
References: <BANLkTimzfrffr2LX7WG00qkSSD-WAHq2Bw@mail.gmail.com>
Message-ID: <BANLkTikH8b9YaJwn8N4W1YH_z86VV27vdg@mail.gmail.com>

Hi Daniel,

Can you try the embedded file and tell me if it works on Snow Leopard ?

Thanks,
Julien.

On Mon, Jun 6, 2011 at 11:06 AM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi devels,
>
> I noticed that the text of the next button is not shown in the
> ctkWorkflowButtonBoxWidget on Snow Leopard.
>
> The underlying ctk code uses
> this->NextButton->setLayoutDirection(Qt::RightToLeft);  to place the
> icon on the right side of the text.
>
> I think that this might be a misuse of the layout direction and causes
> the problems. The back button is set up the same way as the next
> button without the layout direction and works fine.
>
> What do you guys think?
>
> Thanks,
> Daniel
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110606/f32ff1ae/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ctkWorkflowButtonBoxWidget.cpp
Type: text/x-c++src
Size: 18108 bytes
Desc: not available
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110606/f32ff1ae/attachment.cpp>

From s.zelzer at dkfz-heidelberg.de  Tue Jun  7 05:48:42 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Tue, 07 Jun 2011 07:48:42 +0200
Subject: [Ctk-developers] External projects in commontk github organization
Message-ID: <4DEDBBBA.90409@dkfz-heidelberg.de>

Hi guys,

Recently, I observed an explosion in the number of repositories 
containing external repositories in the CommonTK organization of github. 
Could we discuss the need of some of them, and maybe have these 
discussions in the future in advance of pushing those projects?

I am looking especially at the projects below, where Qt might already 
offer enough functionality. Of course I could be wrong, so thank you for 
your feedback!

- sqlite (why not use the Qt sqlite wrapper?)
- KWSys (where is it needed? I did not find any reference to it in CTK)
- libexpat (Qt also provides a streaming XML API, also not used anywhere)
- libcurl (there is also no reference to libcurl in CTK)


Thanks,

Sascha


From jchris.fillionr at kitware.com  Tue Jun  7 13:19:28 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Tue, 7 Jun 2011 09:19:28 -0400
Subject: [Ctk-developers] External projects in commontk github
	organization
In-Reply-To: <4DEDBBBA.90409@dkfz-heidelberg.de>
References: <4DEDBBBA.90409@dkfz-heidelberg.de>
Message-ID: <BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>

Hi Sasha,

You bring a very good point !

The work related to these projects has just been published and is open to
review.
See
https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82

Expat, KWSys, Sqlite and libcurl were initially shipped with MIDAScpp
library ... I advocates to externalize and CMake-ifie properly these
libraries as a first step. That way the code checked in CTK would "really"
corresponds to the library itself.

Having an other perspective is always fruitful :)  You are right, since
MIDAScpp would be part of CTK, it makes sens to not depend on these external
project and rely on Qt directly. Qt is a mandatory dependency after all.

I am cc'ing Zach in the email and I am sure I will be able to provide more
comments.

Zach> Instead of using expat, libcurl, sqlite and kwsys, would it make sens
to depends on Qt ?

In the mean time, I have the following questions:

 - Currently the library is named "CTKMIDAScppCore" and CTKMIDAScppWidgets",
I was thinking we could have something like CTKDataManagementMIDAScppCore
and CTKDataManagementMIDAScppWidgets.
What do you think ? Doing so would leave room to library like
"CTKDataManagementHadoopCore" ...

 - Does introducing a "DataManagement" hierachy of libraries make sens ? As
far as I am concerned, I think it's important to provide both functionality
to process, store and retrieve the data.

Thanks
Jc


On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
<s.zelzer at dkfz-heidelberg.de>wrote:

> Hi guys,
>
> Recently, I observed an explosion in the number of repositories containing
> external repositories in the CommonTK organization of github. Could we
> discuss the need of some of them, and maybe have these discussions in the
> future in advance of pushing those projects?
>
> I am looking especially at the projects below, where Qt might already offer
> enough functionality. Of course I could be wrong, so thank you for your
> feedback!
>
> - sqlite (why not use the Qt sqlite wrapper?)
> - KWSys (where is it needed? I did not find any reference to it in CTK)
> - libexpat (Qt also provides a streaming XML API, also not used anywhere)
> - libcurl (there is also no reference to libcurl in CTK)
>
>
> Thanks,
>
> Sascha
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/a7ad03f6/attachment.htm>

From julien.finet at kitware.com  Tue Jun  7 13:25:49 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Tue, 7 Jun 2011 09:25:49 -0400
Subject: [Ctk-developers] External projects in commontk github
	organization
In-Reply-To: <BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>
References: <4DEDBBBA.90409@dkfz-heidelberg.de>
	<BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>
Message-ID: <BANLkTi=y8YTiFmXMXUuutEC7VzNS-XpFcA@mail.gmail.com>

For the naming convention, is "MIDAScpp" mandatory? couldn't it be just
MIDAS (and be externally refered as CTK MIDAS).
By having MIDAS into CTK, we infer it is C++ and Qt based.
My 2 cts,
Julien.

On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin <
jchris.fillionr at kitware.com> wrote:

> Hi Sasha,
>
> You bring a very good point !
>
> The work related to these projects has just been published and is open to
> review.
> See
> https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>
> Expat, KWSys, Sqlite and libcurl were initially shipped with MIDAScpp
> library ... I advocates to externalize and CMake-ifie properly these
> libraries as a first step. That way the code checked in CTK would "really"
> corresponds to the library itself.
>
> Having an other perspective is always fruitful :)  You are right, since
> MIDAScpp would be part of CTK, it makes sens to not depend on these external
> project and rely on Qt directly. Qt is a mandatory dependency after all.
>
> I am cc'ing Zach in the email and I am sure I will be able to provide more
> comments.
>
> Zach> Instead of using expat, libcurl, sqlite and kwsys, would it make sens
> to depends on Qt ?
>
> In the mean time, I have the following questions:
>
>  - Currently the library is named "CTKMIDAScppCore" and
> CTKMIDAScppWidgets", I was thinking we could have something like
> CTKDataManagementMIDAScppCore and CTKDataManagementMIDAScppWidgets.
> What do you think ? Doing so would leave room to library like
> "CTKDataManagementHadoopCore" ...
>
>  - Does introducing a "DataManagement" hierachy of libraries make sens ? As
> far as I am concerned, I think it's important to provide both functionality
> to process, store and retrieve the data.
>
> Thanks
> Jc
>
>
>
> On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer <s.zelzer at dkfz-heidelberg.de
> > wrote:
>
>> Hi guys,
>>
>> Recently, I observed an explosion in the number of repositories containing
>> external repositories in the CommonTK organization of github. Could we
>> discuss the need of some of them, and maybe have these discussions in the
>> future in advance of pushing those projects?
>>
>> I am looking especially at the projects below, where Qt might already
>> offer enough functionality. Of course I could be wrong, so thank you for
>> your feedback!
>>
>> - sqlite (why not use the Qt sqlite wrapper?)
>> - KWSys (where is it needed? I did not find any reference to it in CTK)
>> - libexpat (Qt also provides a streaming XML API, also not used anywhere)
>> - libcurl (there is also no reference to libcurl in CTK)
>>
>>
>> Thanks,
>>
>> Sascha
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>
>
>
> --
> +1 919 869 8849
>
>
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/dfa54fd1/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Tue Jun  7 14:05:46 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Tue, 07 Jun 2011 16:05:46 +0200
Subject: [Ctk-developers] External projects in commontk github
	organization
In-Reply-To: <BANLkTi=y8YTiFmXMXUuutEC7VzNS-XpFcA@mail.gmail.com>
References: <4DEDBBBA.90409@dkfz-heidelberg.de>
	<BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>
	<BANLkTi=y8YTiFmXMXUuutEC7VzNS-XpFcA@mail.gmail.com>
Message-ID: <4DEE303A.7090306@dkfz-heidelberg.de>

Hi,

naming issues aside, I tend to agree with Julien here.

Generally, I am of course in favor of extending CTK with useful 
functionality. However, here is a short anecdote: Ivo and I recently got 
a request from German group if we think their project would be a good 
fit for CTK. Although it looked very promising, well engineered, and was 
in the scope of CTK, we expressed our concerns because the project was 
not Qt-ified at all and would somehow break the consistent code style we 
have now in CTK. I see the same issue with the current MIDAS code. In 
the past, I would have voted for including MIDAScpp as an external 
project, as we did with all existing projects we wanted to use but did 
not comply with the CTK style.

I do not want to slow people down or to damp their enthusiasm. I am just 
concerned that CTK becomes a Zoo of libraries with inconsistent "Look 
and Feel".

Maybe an open discussion about integrating larger pieces of software *in 
advance* of starting the effort would be helpful in the future.

I am still thinking about your naming suggestions... ;-)

Best,
Sascha

On 06/07/2011 03:25 PM, Julien Finet wrote:
> For the naming convention, is "MIDAScpp" mandatory? couldn't it be 
> just MIDAS (and be externally refered as CTK MIDAS).
> By having MIDAS into CTK, we infer it is C++ and Qt based.
> My 2 cts,
> Julien.
>
> On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin 
> <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>> wrote:
>
>     Hi Sasha,
>
>     You bring a very good point !
>
>     The work related to these projects has just been published and is
>     open to review.
>     See
>     https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>
>     Expat, KWSys, Sqlite and libcurl were initially shipped with
>     MIDAScpp library ... I advocates to externalize and CMake-ifie
>     properly these libraries as a first step. That way the code
>     checked in CTK would "really" corresponds to the library itself.
>
>     Having an other perspective is always fruitful :)  You are right,
>     since MIDAScpp would be part of CTK, it makes sens to not depend
>     on these external project and rely on Qt directly. Qt is a
>     mandatory dependency after all.
>
>     I am cc'ing Zach in the email and I am sure I will be able to
>     provide more comments.
>
>     Zach> Instead of using expat, libcurl, sqlite and kwsys, would it
>     make sens to depends on Qt ?
>
>     In the mean time, I have the following questions:
>
>      - Currently the library is named "CTKMIDAScppCore" and
>     CTKMIDAScppWidgets", I was thinking we could have something like
>     CTKDataManagementMIDAScppCore and CTKDataManagementMIDAScppWidgets.
>     What do you think ? Doing so would leave room to library like
>     "CTKDataManagementHadoopCore" ...
>
>      - Does introducing a "DataManagement" hierachy of libraries make
>     sens ? As far as I am concerned, I think it's important to provide
>     both functionality to process, store and retrieve the data.
>
>     Thanks
>     Jc
>
>
>
>     On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>     <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>>
>     wrote:
>
>         Hi guys,
>
>         Recently, I observed an explosion in the number of
>         repositories containing external repositories in the CommonTK
>         organization of github. Could we discuss the need of some of
>         them, and maybe have these discussions in the future in
>         advance of pushing those projects?
>
>         I am looking especially at the projects below, where Qt might
>         already offer enough functionality. Of course I could be
>         wrong, so thank you for your feedback!
>
>         - sqlite (why not use the Qt sqlite wrapper?)
>         - KWSys (where is it needed? I did not find any reference to
>         it in CTK)
>         - libexpat (Qt also provides a streaming XML API, also not
>         used anywhere)
>         - libcurl (there is also no reference to libcurl in CTK)
>
>
>         Thanks,
>
>         Sascha
>         _______________________________________________
>         Ctk-developers mailing list
>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>         http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
>     -- 
>     +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>
>
>     _______________________________________________
>     Ctk-developers mailing list
>     Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>     http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/bd839d7b/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Tue Jun  7 14:08:49 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Tue, 07 Jun 2011 16:08:49 +0200
Subject: [Ctk-developers] External projects in commontk github
	organization
In-Reply-To: <BANLkTinvGEMWCuzTZKg-vcsibjsUm4Jpgw@mail.gmail.com>
References: <4DEDBBBA.90409@dkfz-heidelberg.de>	<BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>
	<BANLkTinvGEMWCuzTZKg-vcsibjsUm4Jpgw@mail.gmail.com>
Message-ID: <4DEE30F1.4020504@dkfz-heidelberg.de>

Of course Qt depends on a native sqlite implementation, but it provides 
a library which is database-backend independent:

http://doc.qt.nokia.com/4.7/qtsql.html

It uses Qt plugins for delegating the calls to specific RDBMS 
implementations.

Best,
Sascha

On 06/07/2011 03:32 PM, Zach Mullen wrote:
> The MIDAScpp library linked against kwsys/expat/curl because 
> originally it did not depend on Qt.  In the future, we could remove 
> these dependencies; it would just require a lot of rewriting to use 
> the Qt capabilities instead.  Sqlite will have to stay as a dependency 
> since Qt doesn't have any RDBMS capabilities.
>
> -Zach
>
> On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin 
> <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>> wrote:
>
>     Hi Sasha,
>
>     You bring a very good point !
>
>     The work related to these projects has just been published and is
>     open to review.
>     See
>     https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>
>     Expat, KWSys, Sqlite and libcurl were initially shipped with
>     MIDAScpp library ... I advocates to externalize and CMake-ifie
>     properly these libraries as a first step. That way the code
>     checked in CTK would "really" corresponds to the library itself.
>
>     Having an other perspective is always fruitful :)  You are right,
>     since MIDAScpp would be part of CTK, it makes sens to not depend
>     on these external project and rely on Qt directly. Qt is a
>     mandatory dependency after all.
>
>     I am cc'ing Zach in the email and I am sure I will be able to
>     provide more comments.
>
>     Zach> Instead of using expat, libcurl, sqlite and kwsys, would it
>     make sens to depends on Qt ?
>
>     In the mean time, I have the following questions:
>
>      - Currently the library is named "CTKMIDAScppCore" and
>     CTKMIDAScppWidgets", I was thinking we could have something like
>     CTKDataManagementMIDAScppCore and CTKDataManagementMIDAScppWidgets.
>     What do you think ? Doing so would leave room to library like
>     "CTKDataManagementHadoopCore" ...
>
>      - Does introducing a "DataManagement" hierachy of libraries make
>     sens ? As far as I am concerned, I think it's important to provide
>     both functionality to process, store and retrieve the data.
>
>     Thanks
>     Jc
>
>
>     On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>     <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>>
>     wrote:
>
>         Hi guys,
>
>         Recently, I observed an explosion in the number of
>         repositories containing external repositories in the CommonTK
>         organization of github. Could we discuss the need of some of
>         them, and maybe have these discussions in the future in
>         advance of pushing those projects?
>
>         I am looking especially at the projects below, where Qt might
>         already offer enough functionality. Of course I could be
>         wrong, so thank you for your feedback!
>
>         - sqlite (why not use the Qt sqlite wrapper?)
>         - KWSys (where is it needed? I did not find any reference to
>         it in CTK)
>         - libexpat (Qt also provides a streaming XML API, also not
>         used anywhere)
>         - libcurl (there is also no reference to libcurl in CTK)
>
>
>         Thanks,
>
>         Sascha
>         _______________________________________________
>         Ctk-developers mailing list
>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>         http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
>     -- 
>     +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>
>
>
>
> -- 
> Zach Mullen
> R & D Engineer
> Kitware Inc.
> (919) 969-6990 x314
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/5a0cb00f/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Tue Jun  7 14:27:17 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Tue, 07 Jun 2011 16:27:17 +0200
Subject: [Ctk-developers] data exchange in the DICOM Hosting
 implementation / QtSoap Limitation
In-Reply-To: <63D07F9AA364DB4C901F02EF93465C8D780878D789@USMLVA0E934MSX.ww017.siemens.net>
References: <63D07F9AA364DB4C901F02EF93465C8D780878D789@USMLVA0E934MSX.ww017.siemens.net>
Message-ID: <4DEE3545.9030505@dkfz-heidelberg.de>

Hi Michael,

As far as I know, we only tested the data exchange interface with one 
single .dcm file yet. I am also not aware of any restrictions regarding 
the QtSoap array support (apart from a max dimension of 5).

How are you updating the AvailableData structure? Could you send the 
SOAP messages send by the client and received by the server (you might 
have to edit /opt/git/CTK/Plugins/org.commontk.dah.core/ctkSoapLog.h to 
enable logging).

Maybe Ivo and Benoit can also comment on this issue, they where mainly 
involved in writing the data exchange interface.

Thanks,
Sascha

On 06/06/2011 09:35 PM, Caylus, Michael (SCR US) wrote:
> Hi Sascha,
> First thanks a lot for the quick reply I got on my previous post.  I 
> decided to go for the first approach and created my own CTK plugin as 
> we should stay in the context of a hosting application. I managed to 
> start and run my hosted application through the DICOM Application Host 
> interface with this approach.
> I am now looking into passing some dataset to the hosted application.  
> I wanted to start with something simple where the end-user select a 
> folder that contains a list of DICOM files (one series) and pass it to 
> my application. After choosing the folder; I am updating the 
> AvailableData structure and then publish it to the hosted application 
> when I am starting it. By the time the SOAP request is done I could 
> see that the SOAP message will contains my list of uuids I created for 
> one series but when the message is received 
> (ctkSoapMessageProcessorList::process method) on the hosted 
> application; the message is not the same and only contains one uuid 
> now for the series. Is there any limitation currently within QtSoap to 
> support arrays that you are aware of?  I have seen some discussion 
> regarding on the ctk website 
> (_http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment_) 
> but I am not sure what the status here.
> If there is still some restriction with QtSoap regarding the array; 
> should I send multiple notifyDataAvailable messages to the 
> application; one for each file I am willing to get on the application 
> side. Is there any other best alternative beside my first approach?
> Best,
> Michael
> -------------------------------------------------------------------------------------------------------------------------------
> From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de>
> Subject: Re: [Ctk-developers] How to wrap a Qt App into an hosted app
>         confirming to DICOM Supplement 118 / Working Group 23
> To: ctk-developers at commontk.org
> Message-ID: <4DCD447F.7060402 at dkfz-heidelberg.de>
> Content-Type: text/plain; charset="windows-1252"; Format="flowed"
> Hi Michael,
> Welcome to the CTK mailing list.
> DICOM Application Hosting support in CTK is still work in progress, so 
> please be aware that not everything will work yet. Especially the data 
> exchange interface needs more work. Nevertheless, I hope my comments 
> below will get you started to experiment with your own Qt based DICOM 
> App. Comments, bug reports, patches etc. are also always welcome!
> Best,
> Sascha
> On 05/13/2011 03:56 AM, Caylus, Michael (SCR US) wrote:
> > Hello,
> > I am new to the CTK SDK and would like to wrap my Qt application into
> > an hosted app. I have seen such an example by turning on in the CMake
> > configuration:  ctkExampleHost and ctkExampleHosted App.  I came up
> > with the following questions after playing with it:
> >
> >    1. In order to make this wrap-up for my app; I was considering to
> >       edit the org_commontk_dah_exampleapp plug-in which relies on the
> >       org_commontk_dah_app and org_commontk_dah_core plug-in and keep
> >       the ctkExampleHostedApp. Is this approach the right way to
> >       proceed? This implies splitting my app into the hostedApp and a
> >       CTK plugin; with most of the work to be done on the plug-in
> >       level I guess. I have seen some doc explaining how to embed the
> >       CTK plugin framework in my app too. Can I achieve the same
> >       result with this approach and how?
> >
> As you have already seen, it is possible to use the plugin framework 
> from inside your own application and this is also the preferred way 
> (instead of modifying the existing CTK plug-ins directly). In the case 
> of a DICOM App, I can think of two different approaches, depending on 
> your requirements:
> 1.) Your app is a "pure" DICOM App, i.e. you do not need run it 
> outside the context of a "Hosting Application"
> This is the easier approach. You would create your own plug-in, 
> equivalent to org.commontk.dah.exampleapp, which contains all your app 
> code (logic, widgets, etc.). Additionally, you create your own 
> ctkExampleHostedApp executable, which is essentially the same code as 
> in the CTK repository (just adapt paths and plug-in names).
> The
> _http://www.commontk.org/index.php/Documentation/CTK_Plugin_Framework:_Setting_up_a_project?action=purge_
> tutorial explains how to create a CTK-based project.
> 2.) You want to use your app both "stand-alone" and as a "hosted app"
> This involves a little bit more knowledge about the plugin framework, 
> but should also be easy to achieve. What the 
> org.commontk.dah.exampleapp plug-in actually does, is to register a 
> class containing the DICOM "application logic" as a service within the 
> plugin-framework. You could either create this class (similar as the 
> ctkExampleDicomAppLocig class) inside your current project (not in a 
> separate plug-in) and register an instance by using the plugin 
> framework context, or create a small plug-in containing this class and 
> taking care of the registration itself. Then start the plugin 
> framework from inside your executable, if you recognize the host and 
> app url command line parameters (otherwise, create a QMainWindow 
> yourself).
> Depending on how fancy you would like to get concerning the 
> minimization of dependencies of your app in stand-alone mode to the 
> CTK libraries, you could think of a more complex set-up containing two 
> small wrapper executables, one for "app mode" starting the CTK plugin 
> framework and one for "stand-alone mode", both linked to a library 
> containing your application gui and logic...
> >    1. I couldn?t find to which QMainWindow is the button attached to
> >       in the ctkExampleDicomAppLogic::do_something()function.It
> >       doesn?t look like thectkExampleDicomAppLogic is heritating from
> >       a widgetalso. Which object is responsible of the creation of the
> >       mainwindow?
> >
> The org.commontk.dah.exampleapp plugin does not create a QMainWindow. 
> It just create a button an shows it (moved and resized to respect the 
> prescribed screen area from the host). In Qt, widgets become top-level 
> windows if they don't have a parent widget and are "shown".
> >    1. Is there any doc that explains how to create a CTK plug-in and
> >       App from scratch. It would be probably better in my case to
> >       create a new plugin instead of editing an existing one.
> >
> There is no doc about creating a plug-in yet, but you can easily 
> create one by using the ctkPluginGenerator executable (you may have to 
> enable it in CMake). It might have some rough edges though...
> A CTK "app" is actually nothing special. Any executable can do, which 
> links to the CTK libraries. The tutorial linked above explains how to 
> use the plugin framework in an arbitrary executable.
> >    1. How can I add additional includes directories through a
> >       CMakeLists in the CTK framework.  I couldn?t find a way to
> >       define additional directories in the CMakeList of a plug-in (no
> >       include_directories). Is that define somewhere else?
> >
> Inside CTK itself, include directories are set-up auto-magically 
> through our custom CMake dependency system. If you need to use 
> external libraries in your own CTK plug-in (contained in your own 
> project), just use standard CMake INCLUDE_DIRECTORIES (before calling 
> the ctkMacroBuildPlugin macro) and LINK_LIBRARIES(${PROJECT_NAME} ...) 
> (after the ctkMacroBuildPlugin call) commands.
> Happy coding,
> Sascha
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: 
> <_http://public.kitware.com/pipermail/ctk-developers/attachments/20110513/539b8908/attachment-0001.htm_>
> ------------------------------
> _______________________________________________
> Ctk-developers mailing list
> _Ctk-developers at commontk.org_ <mailto:Ctk-developers at commontk.org>
> _http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers_
> End of Ctk-developers Digest, Vol 22, Issue 6
> *********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/8e27e4d7/attachment.htm>

From benoit.bleuze at inria.fr  Tue Jun  7 15:10:28 2011
From: benoit.bleuze at inria.fr (Benoit Bleuze)
Date: Tue, 7 Jun 2011 17:10:28 +0200 (CEST)
Subject: [Ctk-developers] data exchange in the DICOM Hosting
 implementation / QtSoap Limitation
In-Reply-To: <4DEE3545.9030505@dkfz-heidelberg.de>
Message-ID: <1572309921.1274735.1307459428609.JavaMail.root@zmbs4.inria.fr>

Hello Michael, 


First, we always refer to this document when in doubt: http://www.commontk.org/index.php/File:DicomAppHostingSpecs.pdf 
It's been some time since we touched that code, so forgive me if my answers are sometimes a bit vague. 


" I am updating the AvailableData structure and then publish it to the hosted application when I am starting it" 
So I am guessing you are sending a notifyDataAvailable(AvailableData input,true ), with the descriptors for all your files, right? 


Now from your application you should call getData() with the UUIDs you got from the notification, are we still on the same page? 


So If I understood you correctly, the problem is that you only receive on UUID from the notification? 
My question is then: did you use the helper functions to fill in your data structure: 
ctkDicomSoapArrayOfObjectDescriptors ( const QString & name , const ctkDicomAppHosting :: ArrayOfObjectDescriptors & ods ); ArrayOfObjectDescriptor being this: typedef QList < ObjectDescriptor > ArrayOfObjectDescriptors ; 
You must use the types defined in plugins/org.commontk.dah.core/ctkDicomAppHostingTypes.h and the helpers in plugins/org.commontk.dah.core/ctkDicomAppHostingTypesHelpers.h 
You must fill in your AvailableData::patients list, down to the Series level, or just the AvailableData::objectDescriptors array if you just want a quick test (the recommended way is to fill in properly the hierarchy of Patient/study/series properly). 


I think I remember we sent several files at once with Ivo at some point as a test, but I can't be 100% sure. But I am fairly confident this should work. 
Now to unpack the results from soap messages to Qt objects, you must once again use the typeHelpers structures: 
ctkDicomAppHosting :: ArrayOfObjectDescriptors ctkDicomSoapArrayOfObjectDescriptors :: getArrayOfObjectDescriptors ( const QtSoapType & type ) 
Now if this doesn't work we will have to look at the marshalling functions we wrote (all the helper functions). 


Cheers, 
Ben. 


----- Original Message -----


Hi Michael, 

As far as I know, we only tested the data exchange interface with one single .dcm file yet. I am also not aware of any restrictions regarding the QtSoap array support (apart from a max dimension of 5). 

How are you updating the AvailableData structure? Could you send the SOAP messages send by the client and received by the server (you might have to edit /opt/git/CTK/Plugins/org.commontk.dah.core/ctkSoapLog.h to enable logging). 

Maybe Ivo and Benoit can also comment on this issue, they where mainly involved in writing the data exchange interface. 

Thanks, 
Sascha 

On 06/06/2011 09:35 PM, Caylus, Michael (SCR US) wrote: 


Hi Sascha, 

First thanks a lot for the quick reply I got on my previous post. I decided to go for the first approach and created my own CTK plugin as we should stay in the context of a hosting application. I managed to start and run my hosted application through the DICOM Application Host interface with this approach. 

I am now looking into passing some dataset to the hosted application. I wanted to start with something simple where the end-user select a folder that contains a list of DICOM files (one series) and pass it to my application. After choosing the folder; I am updating the AvailableData structure and then publish it to the hosted application when I am starting it. By the time the SOAP request is done I could see that the SOAP message will contains my list of uuids I created for one series but when the message is received ( ctkSoapMessageProcessorList::process method) on the hosted application; the message is not the same and only contains one uuid now for the series. Is there any limitation currently within QtSoap to support arrays that you are aware of? I have seen some discussion regarding on the ctk website ( http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment ) but I am not sure what the status here. 

If there is still some restriction with QtSoap regarding the array; should I send multiple notifyDataAvailable messages to the application; one for each file I am willing to get on the application side. Is there any other best alternative beside my first approach? 

Best, 
Michael 


------------------------------------------------------------------------------------------------------------------------------- 
From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de> 
Subject: Re: [Ctk-developers] How to wrap a Qt App into an hosted app 
confirming to DICOM Supplement 118 / Working Group 23 
To: ctk-developers at commontk.org 
Message-ID: <4DCD447F.7060402 at dkfz-heidelberg.de> 
Content-Type: text/plain; charset="windows-1252"; Format="flowed" 

Hi Michael, 

Welcome to the CTK mailing list. 

DICOM Application Hosting support in CTK is still work in progress, so please be aware that not everything will work yet. Especially the data exchange interface needs more work. Nevertheless, I hope my comments below will get you started to experiment with your own Qt based DICOM App. Comments, bug reports, patches etc. are also always welcome! 

Best, 
Sascha 

On 05/13/2011 03:56 AM, Caylus, Michael (SCR US) wrote: 
> Hello, 
> I am new to the CTK SDK and would like to wrap my Qt application into 
> an hosted app. I have seen such an example by turning on in the CMake 
> configuration: ctkExampleHost and ctkExampleHosted App. I came up 
> with the following questions after playing with it: 
> 
> 1. In order to make this wrap-up for my app; I was considering to 
> edit the org_commontk_dah_exampleapp plug-in which relies on the 
> org_commontk_dah_app and org_commontk_dah_core plug-in and keep 
> the ctkExampleHostedApp. Is this approach the right way to 
> proceed? This implies splitting my app into the hostedApp and a 
> CTK plugin; with most of the work to be done on the plug-in 
> level I guess. I have seen some doc explaining how to embed the 
> CTK plugin framework in my app too. Can I achieve the same 
> result with this approach and how? 
> 
As you have already seen, it is possible to use the plugin framework from inside your own application and this is also the preferred way (instead of modifying the existing CTK plug-ins directly). In the case of a DICOM App, I can think of two different approaches, depending on your requirements: 

1.) Your app is a "pure" DICOM App, i.e. you do not need run it outside the context of a "Hosting Application" 

This is the easier approach. You would create your own plug-in, equivalent to org.commontk.dah.exampleapp, which contains all your app code (logic, widgets, etc.). Additionally, you create your own ctkExampleHostedApp executable, which is essentially the same code as in the CTK repository (just adapt paths and plug-in names). 

The 
http://www.commontk.org/index.php/Documentation/CTK_Plugin_Framework:_Setting_up_a_project?action=purge 
tutorial explains how to create a CTK-based project. 

2.) You want to use your app both "stand-alone" and as a "hosted app" 

This involves a little bit more knowledge about the plugin framework, but should also be easy to achieve. What the org.commontk.dah.exampleapp plug-in actually does, is to register a class containing the DICOM "application logic" as a service within the plugin-framework. You could either create this class (similar as the ctkExampleDicomAppLocig class) inside your current project (not in a separate plug-in) and register an instance by using the plugin framework context, or create a small plug-in containing this class and taking care of the registration itself. Then start the plugin framework from inside your executable, if you recognize the host and app url command line parameters (otherwise, create a QMainWindow yourself). 

Depending on how fancy you would like to get concerning the minimization of dependencies of your app in stand-alone mode to the CTK libraries, you could think of a more complex set-up containing two small wrapper executables, one for "app mode" starting the CTK plugin framework and one for "stand-alone mode", both linked to a library containing your application gui and logic... 

> 1. I couldn?t find to which QMainWindow is the button attached to 
> in the ctkExampleDicomAppLogic::do_something()function.It 
> doesn?t look like thectkExampleDicomAppLogic is heritating from 
> a widgetalso. Which object is responsible of the creation of the 
> mainwindow? 
> 

The org.commontk.dah.exampleapp plugin does not create a QMainWindow. It just create a button an shows it (moved and resized to respect the prescribed screen area from the host). In Qt, widgets become top-level windows if they don't have a parent widget and are "shown". 

> 1. Is there any doc that explains how to create a CTK plug-in and 
> App from scratch. It would be probably better in my case to 
> create a new plugin instead of editing an existing one. 
> 

There is no doc about creating a plug-in yet, but you can easily create one by using the ctkPluginGenerator executable (you may have to enable it in CMake). It might have some rough edges though... 

A CTK "app" is actually nothing special. Any executable can do, which links to the CTK libraries. The tutorial linked above explains how to use the plugin framework in an arbitrary executable. 


> 1. How can I add additional includes directories through a 
> CMakeLists in the CTK framework. I couldn?t find a way to 
> define additional directories in the CMakeList of a plug-in (no 
> include_directories). Is that define somewhere else? 
> 

Inside CTK itself, include directories are set-up auto-magically through our custom CMake dependency system. If you need to use external libraries in your own CTK plug-in (contained in your own project), just use standard CMake INCLUDE_DIRECTORIES (before calling the ctkMacroBuildPlugin macro) and LINK_LIBRARIES(${PROJECT_NAME} ...) (after the ctkMacroBuildPlugin call) commands. 

Happy coding, 

Sascha 
-------------- next part -------------- 
An HTML attachment was scrubbed... 
URL: < http://public.kitware.com/pipermail/ctk-developers/attachments/20110513/539b8908/attachment-0001.htm > 

------------------------------ 

_______________________________________________ 
Ctk-developers mailing list 
Ctk-developers at commontk.org 
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers 


End of Ctk-developers Digest, Vol 22, Issue 6 
********************************************* 



_______________________________________________ 
Ctk-developers mailing list 
Ctk-developers at commontk.org 
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/5da7bbf7/attachment.htm>

From m.nolden at dkfz-heidelberg.de  Tue Jun  7 15:21:15 2011
From: m.nolden at dkfz-heidelberg.de (Marco Nolden)
Date: Tue, 07 Jun 2011 17:21:15 +0200
Subject: [Ctk-developers] External projects in commontk
	github	organization
In-Reply-To: <4DEE30F1.4020504@dkfz-heidelberg.de>
References: <4DEDBBBA.90409@dkfz-heidelberg.de>	<BANLkTimntMzYXme0BiQNDXj7hMd2ga5r+w@mail.gmail.com>	<BANLkTinvGEMWCuzTZKg-vcsibjsUm4Jpgw@mail.gmail.com>
	<4DEE30F1.4020504@dkfz-heidelberg.de>
Message-ID: <4DEE41EB.8020304@dkfz.de>

SQLite is included in the Qt distribution, you can build it integrated
or as a plugin [1]. This is already used in the CTK DICOM classes.

Marco

[1] http://doc.qt.nokia.com/latest/sql-driver.html#qsqlite


On 06/07/2011 04:08 PM, Sascha Zelzer wrote:
> Of course Qt depends on a native sqlite implementation, but it provides
> a library which is database-backend independent:
> 
> http://doc.qt.nokia.com/4.7/qtsql.html
> 
> It uses Qt plugins for delegating the calls to specific RDBMS
> implementations.
> 
> Best,
> Sascha
> 
> On 06/07/2011 03:32 PM, Zach Mullen wrote:
>> The MIDAScpp library linked against kwsys/expat/curl because
>> originally it did not depend on Qt.  In the future, we could remove
>> these dependencies; it would just require a lot of rewriting to use
>> the Qt capabilities instead.  Sqlite will have to stay as a dependency
>> since Qt doesn't have any RDBMS capabilities.
>>
>> -Zach
>>
>> On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin
>> <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>> wrote:
>>
>>     Hi Sasha,
>>
>>     You bring a very good point !
>>
>>     The work related to these projects has just been published and is
>>     open to review.
>>     See
>>     https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>>
>>     Expat, KWSys, Sqlite and libcurl were initially shipped with
>>     MIDAScpp library ... I advocates to externalize and CMake-ifie
>>     properly these libraries as a first step. That way the code
>>     checked in CTK would "really" corresponds to the library itself.
>>
>>     Having an other perspective is always fruitful :)  You are right,
>>     since MIDAScpp would be part of CTK, it makes sens to not depend
>>     on these external project and rely on Qt directly. Qt is a
>>     mandatory dependency after all.
>>
>>     I am cc'ing Zach in the email and I am sure I will be able to
>>     provide more comments.
>>
>>     Zach> Instead of using expat, libcurl, sqlite and kwsys, would it
>>     make sens to depends on Qt ?
>>
>>     In the mean time, I have the following questions:
>>
>>      - Currently the library is named "CTKMIDAScppCore" and
>>     CTKMIDAScppWidgets", I was thinking we could have something like
>>     CTKDataManagementMIDAScppCore and CTKDataManagementMIDAScppWidgets.
>>     What do you think ? Doing so would leave room to library like
>>     "CTKDataManagementHadoopCore" ...
>>
>>      - Does introducing a "DataManagement" hierachy of libraries make
>>     sens ? As far as I am concerned, I think it's important to provide
>>     both functionality to process, store and retrieve the data.
>>
>>     Thanks
>>     Jc
>>
>>
>>     On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>>     <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>>
>>     wrote:
>>
>>         Hi guys,
>>
>>         Recently, I observed an explosion in the number of
>>         repositories containing external repositories in the CommonTK
>>         organization of github. Could we discuss the need of some of
>>         them, and maybe have these discussions in the future in
>>         advance of pushing those projects?
>>
>>         I am looking especially at the projects below, where Qt might
>>         already offer enough functionality. Of course I could be
>>         wrong, so thank you for your feedback!
>>
>>         - sqlite (why not use the Qt sqlite wrapper?)
>>         - KWSys (where is it needed? I did not find any reference to
>>         it in CTK)
>>         - libexpat (Qt also provides a streaming XML API, also not
>>         used anywhere)
>>         - libcurl (there is also no reference to libcurl in CTK)
>>
>>
>>         Thanks,
>>
>>         Sascha
>>         _______________________________________________
>>         Ctk-developers mailing list
>>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>>         http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
>>
>>
>>     -- 
>>     +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>>
>>
>>
>>
>> -- 
>> Zach Mullen
>> R & D Engineer
>> Kitware Inc.
>> (919) 969-6990 x314
>>
> 


-- 
----------------------------------------------------------------------
Dipl.-Inform. Med. Marco Nolden
Deutsches Krebsforschungszentrum       (German Cancer Research Center)
Div. Medical & Biological Informatics          Tel: (+49) 6221-42 2325
Im Neuenheimer Feld 280                        Fax: (+49) 6221-42 2345
D-69120 Heidelberg                             eMail: M.Nolden at dkfz.de


From michael.caylus at siemens.com  Tue Jun  7 19:05:35 2011
From: michael.caylus at siemens.com (Caylus, Michael (SCR US))
Date: Tue, 7 Jun 2011 15:05:35 -0400
Subject: [Ctk-developers] data exchange in the DICOM Hosting
 implementation / QtSoap Limitation
In-Reply-To: <1572309921.1274735.1307459428609.JavaMail.root@zmbs4.inria.fr>
References: <4DEE3545.9030505@dkfz-heidelberg.de>
	<1572309921.1274735.1307459428609.JavaMail.root@zmbs4.inria.fr>
Message-ID: <63D07F9AA364DB4C901F02EF93465C8D780878DCBA@USMLVA0E934MSX.ww017.siemens.net>

Hello Benoit,

Thanks for getting back to me. Please see my comments below.

Cheers,
Michael

From: Benoit Bleuze [mailto:benoit.bleuze at inria.fr]
Sent: Tuesday, June 07, 2011 11:10 AM
To: Sascha Zelzer
Cc: Ljung, Patric (SCR US); ctk-developers at commontk.org; Caylus, Michael (SCR US)
Subject: Re: [Ctk-developers] data exchange in the DICOM Hosting implementation / QtSoap Limitation

Hello Michael,

First, we always refer to this document when in doubt: http://www.commontk.org/index.php/File:DicomAppHostingSpecs.pdf
It's been some time since we touched that code, so forgive me if my answers are sometimes a bit vague.

" I am updating the AvailableData structure and then publish it to the hosted application when I am starting it"
So I am guessing you are sending a notifyDataAvailable(AvailableData input,true ), with the descriptors for all your files, right?

Yes I send just one notification: notifyDataAvailable with all the descriptors for my files.

Now from your application you should call getData() with the UUIDs you got from the notification, are we still on the same page?
Yes that's the plan. I want to collect the all the UUIDs once I received the DataAvailable structure on the application side.

So If I understood you correctly, the problem is that you only receive on UUID from the notification?
Yes I got only one UUID instead of multiple ones and it look like only the last one is picked up.

My question is then: did you use the helper functions to fill in your data structure:
 ctkDicomSoapArrayOfObjectDescriptors(const QString& name,

                               const ctkDicomAppHosting::ArrayOfObjectDescriptors& ods);



ArrayOfObjectDescriptor being this:

typedef QList<ObjectDescriptor> ArrayOfObjectDescriptors;

I am using the helper functions but not the one you mentioned. When I fill the AvailableData structure; it goes down to the series and call this function instead: ctkDicomSoapSeries(const QString& name, const ctkDicomAppHosting::Series& s)

This function will correctly append my list of descriptors for the corresponding series.

When I dump the SOAP message that is sent to the app I can see all my UUIDs for the series but the issue is at the reception when I dump the SOAP message again I only get one UUID (actually the last one).

I have attached to this email the SOAP messages that are being exchanged. The issue seems to me to be more at the QtSoap level than the data helper function.

You must use the types defined in plugins/org.commontk.dah.core/ctkDicomAppHostingTypes.h and the helpers in plugins/org.commontk.dah.core/ctkDicomAppHostingTypesHelpers.h
You must fill in your AvailableData::patients list, down to the Series level,
This is the approach I started with.

or just the AvailableData::objectDescriptors array if you just want a quick test (the recommended way is to fill in properly the hierarchy of Patient/study/series properly).
I will give a try now by sending just the objectDescriptors array now and will let you know.

I think I remember we sent several files at once with Ivo at some point as a test, but I can't be 100% sure. But I am fairly confident this should work.
Now to unpack the results from soap messages to Qt objects, you must once again use the typeHelpers structures:

ctkDicomAppHosting::ArrayOfObjectDescriptors ctkDicomSoapArrayOfObjectDescriptors::getArrayOfObjectDescriptors(const QtSoapType& type)
Now if this doesn't work we will have to look at the marshalling functions we wrote (all the helper functions).

Cheers,
Ben.

________________________________
Hi Michael,

As far as I know, we only tested the data exchange interface with one single .dcm file yet. I am also not aware of any restrictions regarding the QtSoap array support (apart from a max dimension of 5).

How are you updating the AvailableData structure? Could you send the SOAP messages send by the client and received by the server (you might have to edit /opt/git/CTK/Plugins/org.commontk.dah.core/ctkSoapLog.h to enable logging).

Maybe Ivo and Benoit can also comment on this issue, they where mainly involved in writing the data exchange interface.

Thanks,
Sascha

On 06/06/2011 09:35 PM, Caylus, Michael (SCR US) wrote:
Hi Sascha,

First thanks a lot for the quick reply I got on my previous post.  I decided to go for the first approach and created my own CTK plugin as we should stay in the context of a hosting application. I managed to start and run my hosted application through the DICOM Application Host interface with this approach.

I am now looking into passing some dataset to the hosted application.  I wanted to start with something simple where the end-user select a folder that contains a list of DICOM files (one series) and pass it to my application. After choosing the folder; I am updating the AvailableData structure and then publish it to the hosted application when I am starting it. By the time the SOAP request is done I could see that the SOAP message will contains my list of uuids I created for one series but when the message is received (ctkSoapMessageProcessorList::process method) on the hosted application; the message is not the same and only contains one uuid now for the series. Is there any limitation currently within QtSoap to support arrays that you are aware of?  I have seen some discussion regarding on the ctk website (http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment) but I am not sure what the status here.

If there is still some restriction with QtSoap regarding the array; should I send multiple notifyDataAvailable messages to the application; one for each file I am willing to get on the application side. Is there any other best alternative beside my first approach?

Best,
Michael


-------------------------------------------------------------------------------------------------------------------------------
From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de><mailto:s.zelzer at dkfz-heidelberg.de>
Subject: Re: [Ctk-developers] How to wrap a Qt App into an hosted app
        confirming to DICOM Supplement 118 / Working Group 23
To: ctk-developers at commontk.org<mailto:ctk-developers at commontk.org>
Message-ID: <4DCD447F.7060402 at dkfz-heidelberg.de><mailto:4DCD447F.7060402 at dkfz-heidelberg.de>
Content-Type: text/plain; charset="windows-1252"; Format="flowed"

Hi Michael,

Welcome to the CTK mailing list.

DICOM Application Hosting support in CTK is still work in progress, so please be aware that not everything will work yet. Especially the data exchange interface needs more work. Nevertheless, I hope my comments below will get you started to experiment with your own Qt based DICOM App. Comments, bug reports, patches etc. are also always welcome!

Best,
Sascha

On 05/13/2011 03:56 AM, Caylus, Michael (SCR US) wrote:
> Hello,
> I am new to the CTK SDK and would like to wrap my Qt application into
> an hosted app. I have seen such an example by turning on in the CMake
> configuration:  ctkExampleHost and ctkExampleHosted App.  I came up
> with the following questions after playing with it:
>
>    1. In order to make this wrap-up for my app; I was considering to
>       edit the org_commontk_dah_exampleapp plug-in which relies on the
>       org_commontk_dah_app and org_commontk_dah_core plug-in and keep
>       the ctkExampleHostedApp. Is this approach the right way to
>       proceed? This implies splitting my app into the hostedApp and a
>       CTK plugin; with most of the work to be done on the plug-in
>       level I guess. I have seen some doc explaining how to embed the
>       CTK plugin framework in my app too. Can I achieve the same
>       result with this approach and how?
>
As you have already seen, it is possible to use the plugin framework from inside your own application and this is also the preferred way (instead of modifying the existing CTK plug-ins directly). In the case of a DICOM App, I can think of two different approaches, depending on your requirements:

1.) Your app is a "pure" DICOM App, i.e. you do not need run it outside the context of a "Hosting Application"

This is the easier approach. You would create your own plug-in, equivalent to org.commontk.dah.exampleapp, which contains all your app code (logic, widgets, etc.). Additionally, you create your own ctkExampleHostedApp executable, which is essentially the same code as in the CTK repository (just adapt paths and plug-in names).

The
http://www.commontk.org/index.php/Documentation/CTK_Plugin_Framework:_Setting_up_a_project?action=purge
tutorial explains how to create a CTK-based project.

2.) You want to use your app both "stand-alone" and as a "hosted app"

This involves a little bit more knowledge about the plugin framework, but should also be easy to achieve. What the org.commontk.dah.exampleapp plug-in actually does, is to register a class containing the DICOM "application logic" as a service within the plugin-framework. You could either create this class (similar as the ctkExampleDicomAppLocig class) inside your current project (not in a separate plug-in) and register an instance by using the plugin framework context, or create a small plug-in containing this class and taking care of the registration itself. Then start the plugin framework from inside your executable, if you recognize the host and app url command line parameters (otherwise, create a QMainWindow yourself).

Depending on how fancy you would like to get concerning the minimization of dependencies of your app in stand-alone mode to the CTK libraries, you could think of a more complex set-up containing two small wrapper executables, one for "app mode" starting the CTK plugin framework and one for "stand-alone mode", both linked to a library containing your application gui and logic...

>    1. I couldn?t find to which QMainWindow is the button attached to
>       in the ctkExampleDicomAppLogic::do_something()function.It
>       doesn?t look like thectkExampleDicomAppLogic is heritating from
>       a widgetalso. Which object is responsible of the creation of the
>       mainwindow?
>

The org.commontk.dah.exampleapp plugin does not create a QMainWindow. It just create a button an shows it (moved and resized to respect the prescribed screen area from the host). In Qt, widgets become top-level windows if they don't have a parent widget and are "shown".

>    1. Is there any doc that explains how to create a CTK plug-in and
>       App from scratch. It would be probably better in my case to
>       create a new plugin instead of editing an existing one.
>

There is no doc about creating a plug-in yet, but you can easily create one by using the ctkPluginGenerator executable (you may have to enable it in CMake). It might have some rough edges though...

A CTK "app" is actually nothing special. Any executable can do, which links to the CTK libraries. The tutorial linked above explains how to use the plugin framework in an arbitrary executable.


>    1. How can I add additional includes directories through a
>       CMakeLists in the CTK framework.  I couldn?t find a way to
>       define additional directories in the CMakeList of a plug-in (no
>       include_directories). Is that define somewhere else?
>

Inside CTK itself, include directories are set-up auto-magically through our custom CMake dependency system. If you need to use external libraries in your own CTK plug-in (contained in your own project), just use standard CMake INCLUDE_DIRECTORIES (before calling the ctkMacroBuildPlugin macro) and LINK_LIBRARIES(${PROJECT_NAME} ...) (after the ctkMacroBuildPlugin call) commands.

Happy coding,

Sascha
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110513/539b8908/attachment-0001.htm>

------------------------------

_______________________________________________
Ctk-developers mailing list
Ctk-developers at commontk.org<mailto:Ctk-developers at commontk.org>
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers


End of Ctk-developers Digest, Vol 22, Issue 6
*********************************************




_______________________________________________
Ctk-developers mailing list
Ctk-developers at commontk.org
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/265b7d0e/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: soapmessage_host.xml
Type: text/xml
Size: 289635 bytes
Desc: soapmessage_host.xml
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/265b7d0e/attachment.xml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: soapmessage_app.xml
Type: text/xml
Size: 5186 bytes
Desc: soapmessage_app.xml
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/265b7d0e/attachment-0001.xml>

From michael.caylus at siemens.com  Wed Jun  8 05:24:58 2011
From: michael.caylus at siemens.com (Caylus, Michael (SCR US))
Date: Wed, 8 Jun 2011 01:24:58 -0400
Subject: [Ctk-developers] data exchange in the DICOM Hosting
	implementation
In-Reply-To: <mailman.5631.1307456841.4751.ctk-developers@commontk.org>
References: <mailman.5631.1307456841.4751.ctk-developers@commontk.org>
Message-ID: <63D07F9AA364DB4C901F02EF93465C8D780878DE95@USMLVA0E934MSX.ww017.siemens.net>

Hi Sascha,

I have been feeding the available data structure through multiple calls to the addToAvailableData(ctkDicomAppHosting::AvailableData& data,
                        ctkDicomObjectLocatorCache* objectLocatorCache,
                        const QString& filename)

function: one call per file found in my DICOM folder and then I send the notifyDataAvailable message only once.

You should have got the SOAP messages on both the server and client side from my previous mail if not let me know I can send you the files again.

I have also tried to send only the objectDescriptors array as suggested by Ben but I have observed the same issue.

Best,
Michael

-----Original Message-----
From: ctk-developers-bounces at commontk.org [mailto:ctk-developers-bounces at commontk.org] On Behalf Of ctk-developers-request at commontk.org
Sent: Tuesday, June 07, 2011 10:27 AM
To: ctk-developers at commontk.org
Subject: Ctk-developers Digest, Vol 23, Issue 10

Send Ctk-developers mailing list submissions to
        ctk-developers at commontk.org

To subscribe or unsubscribe via the World Wide Web, visit
        http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
or, via email, send a message with subject or body 'help' to
        ctk-developers-request at commontk.org

You can reach the person managing the list at
        ctk-developers-owner at commontk.org

When replying, please edit your Subject line so it is more specific than "Re: Contents of Ctk-developers digest..."


Today's Topics:

   1. Re: External projects in commontk github  organization
      (Sascha Zelzer)
   2. Re: data exchange in the DICOM Hosting implementation /
      QtSoap Limitation (Sascha Zelzer)


----------------------------------------------------------------------

Message: 1
Date: Tue, 07 Jun 2011 16:08:49 +0200
From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de>
Subject: Re: [Ctk-developers] External projects in commontk github
        organization
To: Zach Mullen <zach.mullen at kitware.com>
Cc: "ctk-developers at commontk.org" <ctk-developers at commontk.org>
Message-ID: <4DEE30F1.4020504 at dkfz-heidelberg.de>
Content-Type: text/plain; charset="iso-8859-1"; Format="flowed"

Of course Qt depends on a native sqlite implementation, but it provides a library which is database-backend independent:

http://doc.qt.nokia.com/4.7/qtsql.html

It uses Qt plugins for delegating the calls to specific RDBMS implementations.

Best,
Sascha

On 06/07/2011 03:32 PM, Zach Mullen wrote:
> The MIDAScpp library linked against kwsys/expat/curl because
> originally it did not depend on Qt.  In the future, we could remove
> these dependencies; it would just require a lot of rewriting to use
> the Qt capabilities instead.  Sqlite will have to stay as a dependency
> since Qt doesn't have any RDBMS capabilities.
>
> -Zach
>
> On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin
> <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>> wrote:
>
>     Hi Sasha,
>
>     You bring a very good point !
>
>     The work related to these projects has just been published and is
>     open to review.
>     See
>
> https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a
> 672c9c48c82
>
>     Expat, KWSys, Sqlite and libcurl were initially shipped with
>     MIDAScpp library ... I advocates to externalize and CMake-ifie
>     properly these libraries as a first step. That way the code
>     checked in CTK would "really" corresponds to the library itself.
>
>     Having an other perspective is always fruitful :)  You are right,
>     since MIDAScpp would be part of CTK, it makes sens to not depend
>     on these external project and rely on Qt directly. Qt is a
>     mandatory dependency after all.
>
>     I am cc'ing Zach in the email and I am sure I will be able to
>     provide more comments.
>
>     Zach> Instead of using expat, libcurl, sqlite and kwsys, would it
>     make sens to depends on Qt ?
>
>     In the mean time, I have the following questions:
>
>      - Currently the library is named "CTKMIDAScppCore" and
>     CTKMIDAScppWidgets", I was thinking we could have something like
>     CTKDataManagementMIDAScppCore and CTKDataManagementMIDAScppWidgets.
>     What do you think ? Doing so would leave room to library like
>     "CTKDataManagementHadoopCore" ...
>
>      - Does introducing a "DataManagement" hierachy of libraries make
>     sens ? As far as I am concerned, I think it's important to provide
>     both functionality to process, store and retrieve the data.
>
>     Thanks
>     Jc
>
>
>     On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>     <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>>
>     wrote:
>
>         Hi guys,
>
>         Recently, I observed an explosion in the number of
>         repositories containing external repositories in the CommonTK
>         organization of github. Could we discuss the need of some of
>         them, and maybe have these discussions in the future in
>         advance of pushing those projects?
>
>         I am looking especially at the projects below, where Qt might
>         already offer enough functionality. Of course I could be
>         wrong, so thank you for your feedback!
>
>         - sqlite (why not use the Qt sqlite wrapper?)
>         - KWSys (where is it needed? I did not find any reference to
>         it in CTK)
>         - libexpat (Qt also provides a streaming XML API, also not
>         used anywhere)
>         - libcurl (there is also no reference to libcurl in CTK)
>
>
>         Thanks,
>
>         Sascha
>         _______________________________________________
>         Ctk-developers mailing list
>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
>     --
>     +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>
>
>
>
> --
> Zach Mullen
> R & D Engineer
> Kitware Inc.
> (919) 969-6990 x314
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/5a0cb00f/attachment-0001.htm>

------------------------------

Message: 2
Date: Tue, 07 Jun 2011 16:27:17 +0200
From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de>
Subject: Re: [Ctk-developers] data exchange in the DICOM Hosting
        implementation / QtSoap Limitation
To: "Caylus, Michael (SCR US)" <michael.caylus at siemens.com>
Cc: "Ljung, Patric \(SCR US\)" <patric.ljung at siemens.com>,
        "ctk-developers at commontk.org" <ctk-developers at commontk.org>
Message-ID: <4DEE3545.9030505 at dkfz-heidelberg.de>
Content-Type: text/plain; charset="iso-8859-1"; Format="flowed"

Hi Michael,

As far as I know, we only tested the data exchange interface with one
single .dcm file yet. I am also not aware of any restrictions regarding
the QtSoap array support (apart from a max dimension of 5).

How are you updating the AvailableData structure? Could you send the
SOAP messages send by the client and received by the server (you might
have to edit /opt/git/CTK/Plugins/org.commontk.dah.core/ctkSoapLog.h to
enable logging).

Maybe Ivo and Benoit can also comment on this issue, they where mainly
involved in writing the data exchange interface.

Thanks,
Sascha

On 06/06/2011 09:35 PM, Caylus, Michael (SCR US) wrote:
> Hi Sascha,
> First thanks a lot for the quick reply I got on my previous post.  I
> decided to go for the first approach and created my own CTK plugin as
> we should stay in the context of a hosting application. I managed to
> start and run my hosted application through the DICOM Application Host
> interface with this approach.
> I am now looking into passing some dataset to the hosted application.
> I wanted to start with something simple where the end-user select a
> folder that contains a list of DICOM files (one series) and pass it to
> my application. After choosing the folder; I am updating the
> AvailableData structure and then publish it to the hosted application
> when I am starting it. By the time the SOAP request is done I could
> see that the SOAP message will contains my list of uuids I created for
> one series but when the message is received
> (ctkSoapMessageProcessorList::process method) on the hosted
> application; the message is not the same and only contains one uuid
> now for the series. Is there any limitation currently within QtSoap to
> support arrays that you are aware of?  I have seen some discussion
> regarding on the ctk website
> (_http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment_)
> but I am not sure what the status here.
> If there is still some restriction with QtSoap regarding the array;
> should I send multiple notifyDataAvailable messages to the
> application; one for each file I am willing to get on the application
> side. Is there any other best alternative beside my first approach?
> Best,
> Michael
> -------------------------------------------------------------------------------------------------------------------------------
> From: Sascha Zelzer <s.zelzer at dkfz-heidelberg.de>
> Subject: Re: [Ctk-developers] How to wrap a Qt App into an hosted app
>         confirming to DICOM Supplement 118 / Working Group 23
> To: ctk-developers at commontk.org
> Message-ID: <4DCD447F.7060402 at dkfz-heidelberg.de>
> Content-Type: text/plain; charset="windows-1252"; Format="flowed"
> Hi Michael,
> Welcome to the CTK mailing list.
> DICOM Application Hosting support in CTK is still work in progress, so
> please be aware that not everything will work yet. Especially the data
> exchange interface needs more work. Nevertheless, I hope my comments
> below will get you started to experiment with your own Qt based DICOM
> App. Comments, bug reports, patches etc. are also always welcome!
> Best,
> Sascha
> On 05/13/2011 03:56 AM, Caylus, Michael (SCR US) wrote:
> > Hello,
> > I am new to the CTK SDK and would like to wrap my Qt application into
> > an hosted app. I have seen such an example by turning on in the CMake
> > configuration:  ctkExampleHost and ctkExampleHosted App.  I came up
> > with the following questions after playing with it:
> >
> >    1. In order to make this wrap-up for my app; I was considering to
> >       edit the org_commontk_dah_exampleapp plug-in which relies on the
> >       org_commontk_dah_app and org_commontk_dah_core plug-in and keep
> >       the ctkExampleHostedApp. Is this approach the right way to
> >       proceed? This implies splitting my app into the hostedApp and a
> >       CTK plugin; with most of the work to be done on the plug-in
> >       level I guess. I have seen some doc explaining how to embed the
> >       CTK plugin framework in my app too. Can I achieve the same
> >       result with this approach and how?
> >
> As you have already seen, it is possible to use the plugin framework
> from inside your own application and this is also the preferred way
> (instead of modifying the existing CTK plug-ins directly). In the case
> of a DICOM App, I can think of two different approaches, depending on
> your requirements:
> 1.) Your app is a "pure" DICOM App, i.e. you do not need run it
> outside the context of a "Hosting Application"
> This is the easier approach. You would create your own plug-in,
> equivalent to org.commontk.dah.exampleapp, which contains all your app
> code (logic, widgets, etc.). Additionally, you create your own
> ctkExampleHostedApp executable, which is essentially the same code as
> in the CTK repository (just adapt paths and plug-in names).
> The
> _http://www.commontk.org/index.php/Documentation/CTK_Plugin_Framework:_Setting_up_a_project?action=purge_
> tutorial explains how to create a CTK-based project.
> 2.) You want to use your app both "stand-alone" and as a "hosted app"
> This involves a little bit more knowledge about the plugin framework,
> but should also be easy to achieve. What the
> org.commontk.dah.exampleapp plug-in actually does, is to register a
> class containing the DICOM "application logic" as a service within the
> plugin-framework. You could either create this class (similar as the
> ctkExampleDicomAppLocig class) inside your current project (not in a
> separate plug-in) and register an instance by using the plugin
> framework context, or create a small plug-in containing this class and
> taking care of the registration itself. Then start the plugin
> framework from inside your executable, if you recognize the host and
> app url command line parameters (otherwise, create a QMainWindow
> yourself).
> Depending on how fancy you would like to get concerning the
> minimization of dependencies of your app in stand-alone mode to the
> CTK libraries, you could think of a more complex set-up containing two
> small wrapper executables, one for "app mode" starting the CTK plugin
> framework and one for "stand-alone mode", both linked to a library
> containing your application gui and logic...
> >    1. I couldn?t find to which QMainWindow is the button attached to
> >       in the ctkExampleDicomAppLogic::do_something()function.It
> >       doesn?t look like thectkExampleDicomAppLogic is heritating from
> >       a widgetalso. Which object is responsible of the creation of the
> >       mainwindow?
> >
> The org.commontk.dah.exampleapp plugin does not create a QMainWindow.
> It just create a button an shows it (moved and resized to respect the
> prescribed screen area from the host). In Qt, widgets become top-level
> windows if they don't have a parent widget and are "shown".
> >    1. Is there any doc that explains how to create a CTK plug-in and
> >       App from scratch. It would be probably better in my case to
> >       create a new plugin instead of editing an existing one.
> >
> There is no doc about creating a plug-in yet, but you can easily
> create one by using the ctkPluginGenerator executable (you may have to
> enable it in CMake). It might have some rough edges though...
> A CTK "app" is actually nothing special. Any executable can do, which
> links to the CTK libraries. The tutorial linked above explains how to
> use the plugin framework in an arbitrary executable.
> >    1. How can I add additional includes directories through a
> >       CMakeLists in the CTK framework.  I couldn?t find a way to
> >       define additional directories in the CMakeList of a plug-in (no
> >       include_directories). Is that define somewhere else?
> >
> Inside CTK itself, include directories are set-up auto-magically
> through our custom CMake dependency system. If you need to use
> external libraries in your own CTK plug-in (contained in your own
> project), just use standard CMake INCLUDE_DIRECTORIES (before calling
> the ctkMacroBuildPlugin macro) and LINK_LIBRARIES(${PROJECT_NAME} ...)
> (after the ctkMacroBuildPlugin call) commands.
> Happy coding,
> Sascha
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <_http://public.kitware.com/pipermail/ctk-developers/attachments/20110513/539b8908/attachment-0001.htm_>
> ------------------------------
> _______________________________________________
> Ctk-developers mailing list
> _Ctk-developers at commontk.org_ <mailto:Ctk-developers at commontk.org>
> _http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers_
> End of Ctk-developers Digest, Vol 22, Issue 6
> *********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110607/8e27e4d7/attachment.htm>

------------------------------

_______________________________________________
Ctk-developers mailing list
Ctk-developers at commontk.org
http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers


End of Ctk-developers Digest, Vol 23, Issue 10
**********************************************


From s.zelzer at dkfz-heidelberg.de  Wed Jun  8 09:00:26 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Wed, 08 Jun 2011 11:00:26 +0200
Subject: [Ctk-developers] External projects in commontk github
	organization
Message-ID: <4DEF3A2A.9030104@dkfz-heidelberg.de>

I forward the mail from Zach to ctk-developers, since it does not seem 
to arrive there.

-------- Original Message --------
Subject: 	Re: [Ctk-developers] External projects in commontk github 
organization
Date: 	Tue, 7 Jun 2011 16:35:28 +0200
From: 	Zach Mullen <zach.mullen at kitware.com>
To: 	Zelzer, Sascha <s.zelzer at Dkfz-Heidelberg.de>
CC: 	Julien Finet <julien.finet at kitware.com>, Jean-Christophe 
Fillion-Robin <jchris.fillionr at kitware.com>, 
"ctk-developers at commontk.org" <ctk-developers at commontk.org>



On Tue, Jun 7, 2011 at 10:05 AM, Sascha Zelzer 
<s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:

    Hi,

    naming issues aside, I tend to agree with Julien here.

    Generally, I am of course in favor of extending CTK with useful
    functionality. However, here is a short anecdote: Ivo and I recently
    got a request from German group if we think their project would be a
    good fit for CTK. Although it looked very promising, well
    engineered, and was in the scope of CTK, we expressed our concerns
    because the project was not Qt-ified at all and would somehow break
    the consistent code style we have now in CTK. I see the same issue
    with the current MIDAS code. In the past, I would have voted for
    including MIDAScpp as an external project, as we did with all
    existing projects we wanted to use but did not comply with the CTK
    style.

    I do not want to slow people down or to damp their enthusiasm. I am
    just concerned that CTK becomes a Zoo of libraries with inconsistent
    "Look and Feel".

Hi Sascha,

Part of the original motivation for moving MIDAScpp into CTK was to 
bring it up to the standards of organization and style that CTK uses.  I 
will be heavily maintaining this part of the project, and the effort to 
make MIDAScpp conform to CTK standards will be ongoing.

For now, I will work to remove the KWSys and expat dependencies and 
replace them with Qt functionality before we merge this code.  I will 
also look into replacing curl with the QtNetwork module.

-Zach


    Maybe an open discussion about integrating larger pieces of software
    *in advance* of starting the effort would be helpful in the future.

    I am still thinking about your naming suggestions... ;-)

    Best,
    Sascha


    On 06/07/2011 03:25 PM, Julien Finet wrote:
>     For the naming convention, is "MIDAScpp" mandatory? couldn't it be
>     just MIDAS (and be externally refered as CTK MIDAS).
>     By having MIDAS into CTK, we infer it is C++ and Qt based.
>     My 2 cts,
>     Julien.
>
>     On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin
>     <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>>
>     wrote:
>
>         Hi Sasha,
>
>         You bring a very good point !
>
>         The work related to these projects has just been published and
>         is open to review.
>         See
>         https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>
>         Expat, KWSys, Sqlite and libcurl were initially shipped with
>         MIDAScpp library ... I advocates to externalize and CMake-ifie
>         properly these libraries as a first step. That way the code
>         checked in CTK would "really" corresponds to the library itself.
>
>         Having an other perspective is always fruitful :)  You are
>         right, since MIDAScpp would be part of CTK, it makes sens to
>         not depend on these external project and rely on Qt directly.
>         Qt is a mandatory dependency after all.
>
>         I am cc'ing Zach in the email and I am sure I will be able to
>         provide more comments.
>
>         Zach> Instead of using expat, libcurl, sqlite and kwsys, would
>         it make sens to depends on Qt ?
>
>         In the mean time, I have the following questions:
>
>          - Currently the library is named "CTKMIDAScppCore" and
>         CTKMIDAScppWidgets", I was thinking we could have something
>         like CTKDataManagementMIDAScppCore and
>         CTKDataManagementMIDAScppWidgets.
>         What do you think ? Doing so would leave room to library like
>         "CTKDataManagementHadoopCore" ...
>
>          - Does introducing a "DataManagement" hierachy of libraries
>         make sens ? As far as I am concerned, I think it's important
>         to provide both functionality to process, store and retrieve
>         the data.
>
>         Thanks
>         Jc
>
>
>
>         On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>         <s.zelzer at dkfz-heidelberg.de
>         <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:
>
>             Hi guys,
>
>             Recently, I observed an explosion in the number of
>             repositories containing external repositories in the
>             CommonTK organization of github. Could we discuss the need
>             of some of them, and maybe have these discussions in the
>             future in advance of pushing those projects?
>
>             I am looking especially at the projects below, where Qt
>             might already offer enough functionality. Of course I
>             could be wrong, so thank you for your feedback!
>
>             - sqlite (why not use the Qt sqlite wrapper?)
>             - KWSys (where is it needed? I did not find any reference
>             to it in CTK)
>             - libexpat (Qt also provides a streaming XML API, also not
>             used anywhere)
>             - libcurl (there is also no reference to libcurl in CTK)
>
>
>             Thanks,
>
>             Sascha
>             _______________________________________________
>             Ctk-developers mailing list
>             Ctk-developers at commontk.org
>             <mailto:Ctk-developers at commontk.org>
>             http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
>         -- 
>         +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>
>
>         _______________________________________________
>         Ctk-developers mailing list
>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>         http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110608/0de51c9e/attachment.htm>

From m.nolden at dkfz-heidelberg.de  Wed Jun  8 11:59:45 2011
From: m.nolden at dkfz-heidelberg.de (Marco Nolden)
Date: Wed, 08 Jun 2011 13:59:45 +0200
Subject: [Ctk-developers] External projects in commontk
	github	organization
In-Reply-To: <4DEF3A2A.9030104@dkfz-heidelberg.de>
References: <4DEF3A2A.9030104@dkfz-heidelberg.de>
Message-ID: <4DEF6431.2000709@dkfz.de>

Hi,

I think inside of the CTK repository we should focus on technologies
that are standardized, like we currently do with DICOM Q/R and
application hosting. Having a MIDAS client is probably useful for a lot
of people, but I would prefer to put most of the client implementation
in the MIDAS repository and enable it as an (optional) external project
if we have a use case. Or MIDAS could enable CTK as a "binding" (sorry I
do not know enough about MIDAS) and provide a CTK plugin that could be
used by other CTK-based systems. I think this is the general idea of
having a component based approach and a plugin framework.

Marco








On 06/08/2011 11:00 AM, Sascha Zelzer wrote:
> I forward the mail from Zach to ctk-developers, since it does not seem
> to arrive there.
> 
> -------- Original Message --------
> Subject: 	Re: [Ctk-developers] External projects in commontk github
> organization
> Date: 	Tue, 7 Jun 2011 16:35:28 +0200
> From: 	Zach Mullen <zach.mullen at kitware.com>
> To: 	Zelzer, Sascha <s.zelzer at Dkfz-Heidelberg.de>
> CC: 	Julien Finet <julien.finet at kitware.com>, Jean-Christophe
> Fillion-Robin <jchris.fillionr at kitware.com>,
> "ctk-developers at commontk.org" <ctk-developers at commontk.org>
> 
> 
> 
> On Tue, Jun 7, 2011 at 10:05 AM, Sascha Zelzer
> <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:
> 
>     Hi,
> 
>     naming issues aside, I tend to agree with Julien here.
> 
>     Generally, I am of course in favor of extending CTK with useful
>     functionality. However, here is a short anecdote: Ivo and I recently
>     got a request from German group if we think their project would be a
>     good fit for CTK. Although it looked very promising, well
>     engineered, and was in the scope of CTK, we expressed our concerns
>     because the project was not Qt-ified at all and would somehow break
>     the consistent code style we have now in CTK. I see the same issue
>     with the current MIDAS code. In the past, I would have voted for
>     including MIDAScpp as an external project, as we did with all
>     existing projects we wanted to use but did not comply with the CTK
>     style.
> 
>     I do not want to slow people down or to damp their enthusiasm. I am
>     just concerned that CTK becomes a Zoo of libraries with inconsistent
>     "Look and Feel".
> 
> Hi Sascha,
> 
> Part of the original motivation for moving MIDAScpp into CTK was to
> bring it up to the standards of organization and style that CTK uses.  I
> will be heavily maintaining this part of the project, and the effort to
> make MIDAScpp conform to CTK standards will be ongoing.
> 
> For now, I will work to remove the KWSys and expat dependencies and
> replace them with Qt functionality before we merge this code.  I will
> also look into replacing curl with the QtNetwork module.
> 
> -Zach
> 
> 
>     Maybe an open discussion about integrating larger pieces of software
>     *in advance* of starting the effort would be helpful in the future.
> 
>     I am still thinking about your naming suggestions... ;-)
> 
>     Best,
>     Sascha
> 
> 
>     On 06/07/2011 03:25 PM, Julien Finet wrote:
>>     For the naming convention, is "MIDAScpp" mandatory? couldn't it be
>>     just MIDAS (and be externally refered as CTK MIDAS).
>>     By having MIDAS into CTK, we infer it is C++ and Qt based.
>>     My 2 cts,
>>     Julien.
>>
>>     On Tue, Jun 7, 2011 at 9:19 AM, Jean-Christophe Fillion-Robin
>>     <jchris.fillionr at kitware.com <mailto:jchris.fillionr at kitware.com>>
>>     wrote:
>>
>>         Hi Sasha,
>>
>>         You bring a very good point !
>>
>>         The work related to these projects has just been published and
>>         is open to review.
>>         See
>>         https://github.com/zachmullen/CTK/commit/1051813f5046bb08835b73a12138a672c9c48c82
>>
>>         Expat, KWSys, Sqlite and libcurl were initially shipped with
>>         MIDAScpp library ... I advocates to externalize and CMake-ifie
>>         properly these libraries as a first step. That way the code
>>         checked in CTK would "really" corresponds to the library itself.
>>
>>         Having an other perspective is always fruitful :)  You are
>>         right, since MIDAScpp would be part of CTK, it makes sens to
>>         not depend on these external project and rely on Qt directly.
>>         Qt is a mandatory dependency after all.
>>
>>         I am cc'ing Zach in the email and I am sure I will be able to
>>         provide more comments.
>>
>>         Zach> Instead of using expat, libcurl, sqlite and kwsys, would
>>         it make sens to depends on Qt ?
>>
>>         In the mean time, I have the following questions:
>>
>>          - Currently the library is named "CTKMIDAScppCore" and
>>         CTKMIDAScppWidgets", I was thinking we could have something
>>         like CTKDataManagementMIDAScppCore and
>>         CTKDataManagementMIDAScppWidgets.
>>         What do you think ? Doing so would leave room to library like
>>         "CTKDataManagementHadoopCore" ...
>>
>>          - Does introducing a "DataManagement" hierachy of libraries
>>         make sens ? As far as I am concerned, I think it's important
>>         to provide both functionality to process, store and retrieve
>>         the data.
>>
>>         Thanks
>>         Jc
>>
>>
>>
>>         On Tue, Jun 7, 2011 at 1:48 AM, Sascha Zelzer
>>         <s.zelzer at dkfz-heidelberg.de
>>         <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:
>>
>>             Hi guys,
>>
>>             Recently, I observed an explosion in the number of
>>             repositories containing external repositories in the
>>             CommonTK organization of github. Could we discuss the need
>>             of some of them, and maybe have these discussions in the
>>             future in advance of pushing those projects?
>>
>>             I am looking especially at the projects below, where Qt
>>             might already offer enough functionality. Of course I
>>             could be wrong, so thank you for your feedback!
>>
>>             - sqlite (why not use the Qt sqlite wrapper?)
>>             - KWSys (where is it needed? I did not find any reference
>>             to it in CTK)
>>             - libexpat (Qt also provides a streaming XML API, also not
>>             used anywhere)
>>             - libcurl (there is also no reference to libcurl in CTK)
>>
>>
>>             Thanks,
>>
>>             Sascha
>>             _______________________________________________
>>             Ctk-developers mailing list
>>             Ctk-developers at commontk.org
>>             <mailto:Ctk-developers at commontk.org>
>>             http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
>>
>>
>>         -- 
>>         +1 919 869 8849 <tel:%2B1%20919%20869%208849>
>>
>>
>>         _______________________________________________
>>         Ctk-developers mailing list
>>         Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>>         http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
> 
> 
> 
> 
> 


-- 
----------------------------------------------------------------------
Dipl.-Inform. Med. Marco Nolden
Deutsches Krebsforschungszentrum       (German Cancer Research Center)
Div. Medical & Biological Informatics          Tel: (+49) 6221-42 2325
Im Neuenheimer Feld 280                        Fax: (+49) 6221-42 2345
D-69120 Heidelberg                             eMail: M.Nolden at dkfz.de


From s.zelzer at dkfz-heidelberg.de  Wed Jun  8 15:43:54 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Wed, 08 Jun 2011 17:43:54 +0200
Subject: [Ctk-developers] Mac XCode project errors when building DCMTK
Message-ID: <4DEF98BA.8070208@dkfz-heidelberg.de>

Hi guys,

we have a problem regarding the super-build of DCMTK inside of CKT when 
the XCode CMake generator is used.

The specific problem is in CMakeExternals/DCMTK.cmake:39 where a 
"force-rebuild" step is added. This step is referenced as a dependent 
make target of other targets in the make files generated for the XCode 
project. However, the target does not exist and hence an error is thrown 
by make.

Have other CTK - Mac users experienced the same? Is it a 
"ExternalProject" bug? Or a CMake XCode generator bug?

Many thanks,

Sascha


From jchris.fillionr at kitware.com  Wed Jun  8 16:44:01 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Wed, 8 Jun 2011 12:44:01 -0400
Subject: [Ctk-developers] Mac XCode project errors when building DCMTK
In-Reply-To: <4DEF98BA.8070208@dkfz-heidelberg.de>
References: <4DEF98BA.8070208@dkfz-heidelberg.de>
Message-ID: <BANLkTinp_b3vYsJtU1KKAcTaLgkvtbNr9A@mail.gmail.com>

Hi,

In the mean time, I think we can get rid of the "force-rebuild" step
associated with DCMTK.

This one was added for convenience .. while we were actively working on
DCMTK.

Jc

On Wed, Jun 8, 2011 at 11:43 AM, Sascha Zelzer
<s.zelzer at dkfz-heidelberg.de>wrote:

> Hi guys,
>
> we have a problem regarding the super-build of DCMTK inside of CKT when the
> XCode CMake generator is used.
>
> The specific problem is in CMakeExternals/DCMTK.cmake:39 where a
> "force-rebuild" step is added. This step is referenced as a dependent make
> target of other targets in the make files generated for the XCode project.
> However, the target does not exist and hence an error is thrown by make.
>
> Have other CTK - Mac users experienced the same? Is it a "ExternalProject"
> bug? Or a CMake XCode generator bug?
>
> Many thanks,
>
> Sascha
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110608/6648ceb5/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Wed Jun  8 16:48:34 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Wed, 08 Jun 2011 18:48:34 +0200
Subject: [Ctk-developers] data exchange in the DICOM
	Hosting	implementation
In-Reply-To: <63D07F9AA364DB4C901F02EF93465C8D780878DE95@USMLVA0E934MSX.ww017.siemens.net>
References: <mailman.5631.1307456841.4751.ctk-developers@commontk.org>
	<63D07F9AA364DB4C901F02EF93465C8D780878DE95@USMLVA0E934MSX.ww017.siemens.net>
Message-ID: <4DEFA7E2.3080809@dkfz-heidelberg.de>

Hi Michael,

it looks like the deserialization code of QtSoapArray is buggy. When 
serializing a QtSoapArray instance to XML, it writes a position 
attribute like "[3]", but when reading it in, this string is getting 
converted to an integer which always leads to 0 (because the conversion 
fails due to the brackets). Please try to replace line 1284 in 
qtsoap.cpp (located in your CTK superbuild build tree) with:

pos = posattr.value().mid(1, posattr.value().size()-2).toInt();

and make a build at the superbuild level. If that solves your problem I 
will push that fix to our repo and update
http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment 
with a comment. I am aware of the limitation that this will only allow 
to deserialize one-dimensional arrays, but I think there is no 
multi-dimensional use case in the DICOM AppHosting standard, so I would 
refrain from writing support for it until we need it.

Best,
Sascha

On 06/08/2011 07:24 AM, Caylus, Michael (SCR US) wrote:
> Hi Sascha,
>
> I have been feeding the available data structure through multiple calls to the addToAvailableData(ctkDicomAppHosting::AvailableData&  data,
>                          ctkDicomObjectLocatorCache* objectLocatorCache,
>                          const QString&  filename)
>
> function: one call per file found in my DICOM folder and then I send the notifyDataAvailable message only once.
>
> You should have got the SOAP messages on both the server and client side from my previous mail if not let me know I can send you the files again.
>
> I have also tried to send only the objectDescriptors array as suggested by Ben but I have observed the same issue.
>
> Best,
> Michael
>



From michael.caylus at siemens.com  Thu Jun  9 01:16:55 2011
From: michael.caylus at siemens.com (Caylus, Michael (SCR US))
Date: Wed, 8 Jun 2011 21:16:55 -0400
Subject: [Ctk-developers] data exchange in the DICOM
	Hosting	implementation
In-Reply-To: <4DEFA7E2.3080809@dkfz-heidelberg.de>
References: <mailman.5631.1307456841.4751.ctk-developers@commontk.org>
	<63D07F9AA364DB4C901F02EF93465C8D780878DE95@USMLVA0E934MSX.ww017.siemens.net>
	<4DEFA7E2.3080809@dkfz-heidelberg.de>
Message-ID: <63D07F9AA364DB4C901F02EF93465C8D78089A1D1F@USMLVA0E934MSX.ww017.siemens.net>

Hi Sascha,

You can push that fix to your repo. The fix you provided fixed my issue.

Thanks for your help!

Best,
Michael

-----Original Message-----
From: Sascha Zelzer [mailto:s.zelzer at dkfz-heidelberg.de] 
Sent: Wednesday, June 08, 2011 12:49 PM
To: Caylus, Michael (SCR US)
Cc: ctk-developers at commontk.org; Ljung, Patric (SCR US)
Subject: Re: [Ctk-developers] data exchange in the DICOM Hosting implementation

Hi Michael,

it looks like the deserialization code of QtSoapArray is buggy. When serializing a QtSoapArray instance to XML, it writes a position attribute like "[3]", but when reading it in, this string is getting converted to an integer which always leads to 0 (because the conversion fails due to the brackets). Please try to replace line 1284 in qtsoap.cpp (located in your CTK superbuild build tree) with:

pos = posattr.value().mid(1, posattr.value().size()-2).toInt();

and make a build at the superbuild level. If that solves your problem I will push that fix to our repo and update http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment
with a comment. I am aware of the limitation that this will only allow to deserialize one-dimensional arrays, but I think there is no multi-dimensional use case in the DICOM AppHosting standard, so I would refrain from writing support for it until we need it.

Best,
Sascha

On 06/08/2011 07:24 AM, Caylus, Michael (SCR US) wrote:
> Hi Sascha,
>
> I have been feeding the available data structure through multiple calls to the addToAvailableData(ctkDicomAppHosting::AvailableData&  data,
>                          ctkDicomObjectLocatorCache* objectLocatorCache,
>                          const QString&  filename)
>
> function: one call per file found in my DICOM folder and then I send the notifyDataAvailable message only once.
>
> You should have got the SOAP messages on both the server and client side from my previous mail if not let me know I can send you the files again.
>
> I have also tried to send only the objectDescriptors array as suggested by Ben but I have observed the same issue.
>
> Best,
> Michael
>



From benoit.bleuze at inria.fr  Thu Jun  9 07:49:22 2011
From: benoit.bleuze at inria.fr (Benoit Bleuze)
Date: Thu, 9 Jun 2011 09:49:22 +0200 (CEST)
Subject: [Ctk-developers] data exchange in the
	DICOM	Hosting	implementation
In-Reply-To: <63D07F9AA364DB4C901F02EF93465C8D78089A1D1F@USMLVA0E934MSX.ww017.siemens.net>
Message-ID: <1057858584.1301036.1307605762486.JavaMail.root@zmbs4.inria.fr>

Good news for all of us!

Indeed Sascha, there are only 1-dimensional arrays in the specs.
By the way do you know how active is the development of Qt Soap? I had the impression it was quite dormant, wasn't it?

Cheers,
Ben


----- Original Message -----
> Hi Sascha,
> 
> You can push that fix to your repo. The fix you provided fixed my
> issue.
> 
> Thanks for your help!
> 
> Best,
> Michael
> 
> -----Original Message-----
> From: Sascha Zelzer [mailto:s.zelzer at dkfz-heidelberg.de]
> Sent: Wednesday, June 08, 2011 12:49 PM
> To: Caylus, Michael (SCR US)
> Cc: ctk-developers at commontk.org; Ljung, Patric (SCR US)
> Subject: Re: [Ctk-developers] data exchange in the DICOM Hosting
> implementation
> 
> Hi Michael,
> 
> it looks like the deserialization code of QtSoapArray is buggy. When
> serializing a QtSoapArray instance to XML, it writes a position
> attribute like "[3]", but when reading it in, this string is getting
> converted to an integer which always leads to 0 (because the
> conversion fails due to the brackets). Please try to replace line 1284
> in qtsoap.cpp (located in your CTK superbuild build tree) with:
> 
> pos = posattr.value().mid(1, posattr.value().size()-2).toInt();
> 
> and make a build at the superbuild level. If that solves your problem
> I will push that fix to our repo and update
> http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment
> with a comment. I am aware of the limitation that this will only allow
> to deserialize one-dimensional arrays, but I think there is no
> multi-dimensional use case in the DICOM AppHosting standard, so I
> would refrain from writing support for it until we need it.
> 
> Best,
> Sascha
> 
> On 06/08/2011 07:24 AM, Caylus, Michael (SCR US) wrote:
> > Hi Sascha,
> >
> > I have been feeding the available data structure through multiple
> > calls to the addToAvailableData(ctkDicomAppHosting::AvailableData&
> > data,
> >                          ctkDicomObjectLocatorCache*
> >                          objectLocatorCache,
> >                          const QString& filename)
> >
> > function: one call per file found in my DICOM folder and then I send
> > the notifyDataAvailable message only once.
> >
> > You should have got the SOAP messages on both the server and client
> > side from my previous mail if not let me know I can send you the
> > files again.
> >
> > I have also tried to send only the objectDescriptors array as
> > suggested by Ben but I have observed the same issue.
> >
> > Best,
> > Michael
> >
> 
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers


From s.zelzer at dkfz-heidelberg.de  Thu Jun  9 12:06:38 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Thu, 09 Jun 2011 14:06:38 +0200
Subject: [Ctk-developers] data exchange in
	the	DICOM	Hosting	implementation
In-Reply-To: <1057858584.1301036.1307605762486.JavaMail.root@zmbs4.inria.fr>
References: <1057858584.1301036.1307605762486.JavaMail.root@zmbs4.inria.fr>
Message-ID: <4DF0B74E.1050706@dkfz-heidelberg.de>

Yes, AFAIK the development of QtSOAP is "complete" in a sense that is 
does the things it was designed for (except for having some bugs, 
obviously...). I think Nokia just makes sure that it compiles with newer 
Qt versions.

Cheers,
Sascha

On 06/09/2011 09:49 AM, Benoit Bleuze wrote:
> Good news for all of us!
>
> Indeed Sascha, there are only 1-dimensional arrays in the specs.
> By the way do you know how active is the development of Qt Soap? I had the impression it was quite dormant, wasn't it?
>
> Cheers,
> Ben
>
>
> ----- Original Message -----
>> Hi Sascha,
>>
>> You can push that fix to your repo. The fix you provided fixed my
>> issue.
>>
>> Thanks for your help!
>>
>> Best,
>> Michael
>>
>> -----Original Message-----
>> From: Sascha Zelzer [mailto:s.zelzer at dkfz-heidelberg.de]
>> Sent: Wednesday, June 08, 2011 12:49 PM
>> To: Caylus, Michael (SCR US)
>> Cc: ctk-developers at commontk.org; Ljung, Patric (SCR US)
>> Subject: Re: [Ctk-developers] data exchange in the DICOM Hosting
>> implementation
>>
>> Hi Michael,
>>
>> it looks like the deserialization code of QtSoapArray is buggy. When
>> serializing a QtSoapArray instance to XML, it writes a position
>> attribute like "[3]", but when reading it in, this string is getting
>> converted to an integer which always leads to 0 (because the
>> conversion fails due to the brackets). Please try to replace line 1284
>> in qtsoap.cpp (located in your CTK superbuild build tree) with:
>>
>> pos = posattr.value().mid(1, posattr.value().size()-2).toInt();
>>
>> and make a build at the superbuild level. If that solves your problem
>> I will push that fix to our repo and update
>> http://www.commontk.org/index.php/Documentation/DicomApplicationHosting:DataExchangeDevelopment
>> with a comment. I am aware of the limitation that this will only allow
>> to deserialize one-dimensional arrays, but I think there is no
>> multi-dimensional use case in the DICOM AppHosting standard, so I
>> would refrain from writing support for it until we need it.
>>
>> Best,
>> Sascha
>>
>> On 06/08/2011 07:24 AM, Caylus, Michael (SCR US) wrote:
>>> Hi Sascha,
>>>
>>> I have been feeding the available data structure through multiple
>>> calls to the addToAvailableData(ctkDicomAppHosting::AvailableData&
>>> data,
>>>                           ctkDicomObjectLocatorCache*
>>>                           objectLocatorCache,
>>>                           const QString&  filename)
>>>
>>> function: one call per file found in my DICOM folder and then I send
>>> the notifyDataAvailable message only once.
>>>
>>> You should have got the SOAP messages on both the server and client
>>> side from my previous mail if not let me know I can send you the
>>> files again.
>>>
>>> I have also tried to send only the objectDescriptors array as
>>> suggested by Ben but I have observed the same issue.
>>>
>>> Best,
>>> Michael
>>>
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers



From s.zelzer at dkfz-heidelberg.de  Thu Jun  9 12:09:29 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Thu, 09 Jun 2011 14:09:29 +0200
Subject: [Ctk-developers] Mac XCode project errors when building DCMTK
In-Reply-To: <BANLkTinp_b3vYsJtU1KKAcTaLgkvtbNr9A@mail.gmail.com>
References: <4DEF98BA.8070208@dkfz-heidelberg.de>
	<BANLkTinp_b3vYsJtU1KKAcTaLgkvtbNr9A@mail.gmail.com>
Message-ID: <4DF0B7F9.2020706@dkfz-heidelberg.de>

Hi Jc,

thanks for the info. I would then disable the step in the CTK - DCMTK.

Do you think it makes sense to file a bug report for the XCode CMake 
generator?

Thanks,
Sascha

On 06/08/2011 06:44 PM, Jean-Christophe Fillion-Robin wrote:
> Hi,
>
> In the mean time, I think we can get rid of the "force-rebuild" step 
> associated with DCMTK.
>
> This one was added for convenience .. while we were actively working 
> on DCMTK.
>
> Jc
>
> On Wed, Jun 8, 2011 at 11:43 AM, Sascha Zelzer 
> <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:
>
>     Hi guys,
>
>     we have a problem regarding the super-build of DCMTK inside of CKT
>     when the XCode CMake generator is used.
>
>     The specific problem is in CMakeExternals/DCMTK.cmake:39 where a
>     "force-rebuild" step is added. This step is referenced as a
>     dependent make target of other targets in the make files generated
>     for the XCode project. However, the target does not exist and
>     hence an error is thrown by make.
>
>     Have other CTK - Mac users experienced the same? Is it a
>     "ExternalProject" bug? Or a CMake XCode generator bug?
>
>     Many thanks,
>
>     Sascha
>     _______________________________________________
>     Ctk-developers mailing list
>     Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>     http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
> -- 
> +1 919 869 8849
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110609/5f3d940a/attachment.htm>

From julien.finet at kitware.com  Thu Jun  9 17:16:39 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Thu, 9 Jun 2011 13:16:39 -0400
Subject: [Ctk-developers] ctkDICOMImage in ctkDICOMDatabase
Message-ID: <BANLkTikQwAs0jG3M66A-nHq6bc5a=jwzMg@mail.gmail.com>

Hi Marco/Steve,

ctkDICOMDatabase is in Libs/DICOM/Core which means it shouldn't depend on
QtGUI, however, it internally uses ctkDICOMImage which is QtGUI depend (and
needs to be moved in Libs/DICOM/Widgets).

Is ctkDICOMImage necessary in ctkDICOMDatabase? Right now, it seems to be
just a test. Are you ok if I remove it ?

Thanks,
Julien.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110609/ac65cc5f/attachment.htm>

From pieper at ibility.net  Thu Jun  9 17:44:02 2011
From: pieper at ibility.net (Steve Pieper)
Date: Thu, 9 Jun 2011 13:44:02 -0400
Subject: [Ctk-developers] ctkDICOMImage in ctkDICOMDatabase
In-Reply-To: <BANLkTikQwAs0jG3M66A-nHq6bc5a=jwzMg@mail.gmail.com>
References: <BANLkTikQwAs0jG3M66A-nHq6bc5a=jwzMg@mail.gmail.com>
Message-ID: <BANLkTi=OS1dUcMSBpovpW4L_NzsVA+mC0g@mail.gmail.com>

Hi Julien -

The idea here is that the ctkDICOMDatabase should be able to
automatically make the thumbnail images (png) on the fly as dicom
files are added.  ctkDICOMImage uses the QImage class to do this,
which is in QtGui even though it seems to me to be a purely data
formatting class, not a display class.  Even though it's still being
debugged, I'm planning to make use of the thumbnail feature in
slicer4's use of ctkDICOM.

Can you think of a way we can encapsulate the thumbnail calculation in
ctkDICOMCore?  Perhaps we should make an abstract thumbnail generator
class in ctkDICOMCore and then have a QImage-based implementation in
ctkDICOMWidgets that is registered with ctkDICOMDatabase at run time.
Alternatively perhaps ctkDICOMDatabase could link directly to the png
library and make the thumbnails with that.

-Steve


On Thu, Jun 9, 2011 at 1:16 PM, Julien Finet <julien.finet at kitware.com> wrote:
> Hi Marco/Steve,
> ctkDICOMDatabase is in Libs/DICOM/Core which means it shouldn't depend on
> QtGUI, however, it internally uses ctkDICOMImage which is QtGUI depend (and
> needs to be moved in Libs/DICOM/Widgets).
> Is ctkDICOMImage necessary in ctkDICOMDatabase? Right now, it seems to be
> just a test. Are you ok if I remove it ?
> Thanks,
> Julien.


From julien.finet at kitware.com  Thu Jun  9 17:57:02 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Thu, 9 Jun 2011 13:57:02 -0400
Subject: [Ctk-developers] ctkDICOMImage in ctkDICOMDatabase
In-Reply-To: <BANLkTi=OS1dUcMSBpovpW4L_NzsVA+mC0g@mail.gmail.com>
References: <BANLkTikQwAs0jG3M66A-nHq6bc5a=jwzMg@mail.gmail.com>
	<BANLkTi=OS1dUcMSBpovpW4L_NzsVA+mC0g@mail.gmail.com>
Message-ID: <BANLkTikjwSVVS4_p1d3xY6DK2UQ_HYsqNQ@mail.gmail.com>

It seems that if ctkDICOMDatabase is in CTKCore, then it doesn't need
thumbnails.
Maybe ctkDICOMDatabase could be overloaded in CTKWidget and handles
thumbnails there.
Julien.

On Thu, Jun 9, 2011 at 1:44 PM, Steve Pieper <pieper at ibility.net> wrote:

> Hi Julien -
>
> The idea here is that the ctkDICOMDatabase should be able to
> automatically make the thumbnail images (png) on the fly as dicom
> files are added.  ctkDICOMImage uses the QImage class to do this,
> which is in QtGui even though it seems to me to be a purely data
> formatting class, not a display class.  Even though it's still being
> debugged, I'm planning to make use of the thumbnail feature in
> slicer4's use of ctkDICOM.
>
> Can you think of a way we can encapsulate the thumbnail calculation in
> ctkDICOMCore?  Perhaps we should make an abstract thumbnail generator
> class in ctkDICOMCore and then have a QImage-based implementation in
> ctkDICOMWidgets that is registered with ctkDICOMDatabase at run time.
> Alternatively perhaps ctkDICOMDatabase could link directly to the png
> library and make the thumbnails with that.
>
> -Steve
>
>
> On Thu, Jun 9, 2011 at 1:16 PM, Julien Finet <julien.finet at kitware.com>
> wrote:
> > Hi Marco/Steve,
> > ctkDICOMDatabase is in Libs/DICOM/Core which means it shouldn't depend on
> > QtGUI, however, it internally uses ctkDICOMImage which is QtGUI depend
> (and
> > needs to be moved in Libs/DICOM/Widgets).
> > Is ctkDICOMImage necessary in ctkDICOMDatabase? Right now, it seems to be
> > just a test. Are you ok if I remove it ?
> > Thanks,
> > Julien.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110609/16fd7ee6/attachment.htm>

From michael.caylus at siemens.com  Tue Jun 14 21:43:23 2011
From: michael.caylus at siemens.com (Caylus, Michael (SCR US))
Date: Tue, 14 Jun 2011 17:43:23 -0400
Subject: [Ctk-developers] Clarification data handling - Application Hosting
	Support 118
Message-ID: <63D07F9AA364DB4C901F02EF93465C8D7808A545F6@USMLVA0E934MSX.ww017.siemens.net>

Hi all,

I have some uncertainties related to the data handling support in the DICOM Support 118 specification:

In the file-based model in Figure 8.3.1 of spec; we can see that the selection of Data is done though the Host. It is then the responsibility of the application to pull information from the host about the input data.

1. Is it required to have any user interaction on the application to request the data we want to retrieve? Can it happen automatically right after processing the call of the notifyDataAvailable message ? In the current example of the DICOM 118 support implemented in the CTK SDK; I could see that the request of data is initiated from the user through a mouse click on the button created by the application. I am looking into a way to send a getData message without user interaction so that basically the application shows directly the available data selected on the host without any user interaction on the application.  I have tried to make a call to getData within the notifyDataAvailable (ctkExampleDicomAppLogic) function but it looks like first the notifyDataAvailable message need to be processed before making a getData request.

2. This leads me to a second question: Where should occur the selection of Data? On the Host where the user define all the available data and on the  Application where we want to define which input Data need to be retrieved depending of the available data? My first intuition on this was that selection should only be done on the Host but then I guess an extra notification is required so that the application  knows that the notifyDataAvailable message has been fully process.

Best,
Michael

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110614/bda5e373/attachment.htm>

From m.nolden at dkfz-heidelberg.de  Fri Jun 17 14:47:52 2011
From: m.nolden at dkfz-heidelberg.de (Marco Nolden)
Date: Fri, 17 Jun 2011 16:47:52 +0200
Subject: [Ctk-developers] CTK and DCMTK 3.6 compatibility
Message-ID: <4DFB6918.6060304@dkfz.de>

Hi,

I started using the CTK DICOM Widgets in MITK. Since we want to use the
official 3.6 release of DCMTK there are some compatibility issues. To
get started I fixed the constant names, adapted the superbuild and put
this in a new branch

https://github.com/commontk/CTK/commits/dcmtk-3.6

As far as I can see there are no build issues with DCMTK itself, the
stock CMake files of DCMTK 3.6 seem to work fine.
However the interface of the DcmSCU class has changed and I'm not sure
how to replace the methods correctly and whether the implementation in
3.6 is already complete.

Best,

Marco


From jchris.fillionr at kitware.com  Fri Jun 17 16:25:00 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Fri, 17 Jun 2011 12:25:00 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTi=sYr6megrCtJR3b2nE2Tj4mpsdTw@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
	<BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>
	<BANLkTimJ58=syYEvZqNc7YNT7G0xNZBg_A@mail.gmail.com>
	<BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
	<BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>
	<BANLkTinMLEtQvjRcOuq-f9VaV6NvCnteMw@mail.gmail.com>
	<BANLkTi=sYr6megrCtJR3b2nE2Tj4mpsdTw@mail.gmail.com>
Message-ID: <BANLkTi=sPbsoUkJ2tSAoU_uQ+aN2dygLPw@mail.gmail.com>

Hi Daniel,

Can we consider the case closed ?

Could you document the issue on the bugtracker (label: Feature Request) ?
See https://github.com/commontk/CTK/issues

Thanks
Jc

On Sat, Jun 4, 2011 at 10:41 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:

> Hi Danielle,
>
> that is absolutely correct!
>
> Maybe it would be easiest to just fire a
> d->workflow->currentStepChanged(this) signal when
> ctkWorkflowStep::setName or setDescription are called? Then, the
> groupbox would be automatically updated.
>
> Nevertheless, it has no high priority for me since I now set the name
> and description of the step in its constructor - before adding it to
> the workflow.
>
> Cheers,
> Daniel
>
> On Fri, Jun 3, 2011 at 4:13 PM, Danielle Pace <danielle.pace at kitware.com>
> wrote:
> > Hi Daniel,
> > Actually, I think you exposed something important here.
> > I'm assuming, from python, that you wrote something similar to:
> > - create step 1 (with no name/description)
> > - create step 2 (with no name/description)
> > - create workflow widget and add steps
> > - display workflow widget
> > - set name/description for steps 1 and 2
> > - expected to see workflow update, and it didn't
> > Is this correct?  I think we are missing some signals on ctkWorkflowStep
> > such as NameChanged(QString) and DescriptionChanged(QString) that the
> > ctkWorkflowGroupBox should be listening to.
> > Perhaps you could still post your snippet of python code that wasn't
> > working, so I can verify that it is an ordering of commands that we
> should
> > support?
> > Thanks,
> >
> > Danielle
> >
> >
> >
> > On Fri, Jun 3, 2011 at 3:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
> wrote:
> >>
> >> Hi Danielle,
> >>
> >> thanks to your response I checked it again and you are right: it was
> >> my mistake :)
> >>
> >> If the name and description of the steps are set before the steps are
> >> added to the workflow, it all works as expected.
> >>
> >> I set the stuff after adding it and of course it did not recognize it
> >> until the signal was fired.
> >>
> >> Thanks for the great work!
> >>
> >> Cheers,
> >> Daniel
> >>
> >> On Fri, Jun 3, 2011 at 3:44 PM, Danielle Pace <
> danielle.pace at kitware.com>
> >> wrote:
> >> > Hi Daniel,
> >> > You shouldn't have to call the currenStepChanged signal directly.  Can
> >> > you
> >> > please post your bit of code that is not working to the CTK developers
> >> > mailing list, and I will take a look?
> >> > Thanks,
> >> >
> >> > Danielle
> >> >
> >> > On Fri, Jun 3, 2011 at 3:21 PM, Julien Finet <
> julien.finet at kitware.com>
> >> > wrote:
> >> >>
> >> >>
> >> >> ---------- Forwarded message ----------
> >> >> From: Daniel Haehn <haehn at bwh.harvard.edu>
> >> >> Date: Fri, Jun 3, 2011 at 3:13 PM
> >> >> Subject: Re: [Ctk-developers] Setting name and description of
> >> >> ctkWorkflowWidgetStep
> >> >> To: ctk-developers at commontk.org
> >> >>
> >> >>
> >> >> Hi guys,
> >> >>
> >> >> I got it:
> >> >>
> >> >> the ctkWorkflowGroupBox listens for the currentStepChanged signal to
> >> >> update the title (==name) and subTitle (==description) for the
> current
> >> >> step.
> >> >>
> >> >> So maybe it makes sense to invoke this signal when the
> >> >> ctkWorkflow.start() gets called?
> >> >>
> >> >> If I do it manually like that, it works:
> >> >>
> >> >>      self.workflow.start()
> >> >>      self.workflow.currentStepChanged(steps[0])
> >> >>
> >> >> Cheers,
> >> >> Daniel
> >> >>
> >> >> On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn <haehn at bwh.harvard.edu
> >
> >> >> wrote:
> >> >> > Hi devels,
> >> >> >
> >> >> > in C++ I can do
> >> >> >
> >> >> >  this->setName("Title of step");
> >> >> >  this->setDescription("Instructions");
> >> >> >
> >> >> > in a ctkWorkflowWidgetStep, which then get nicely displayed in the
> >> >> > associated ctkWorkflowGroupBox.
> >> >> >
> >> >> > In Python, the same methods can be accessed but the
> >> >> > ctkWorkflowGroupBox does not seem to show them.
> >> >> >
> >> >> > Do I need to trigger an update on the box or something similar?
> >> >> >
> >> >> > Thank you!
> >> >> > Daniel
> >> >> >
> >> >> _______________________________________________
> >> >> Ctk-developers mailing list
> >> >> Ctk-developers at commontk.org
> >> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
> >> >>
> >> >
> >> >
> >> >
> >> > --
> >> > Danielle Pace, M.ESc.
> >> > Research and Development Engineer
> >> > Kitware Inc.,
> >> > North Carolina Office
> >> > www.kitware.com
> >> > 919-969-6990 X 319
> >> >
> >
> >
> >
> > --
> > Danielle Pace, M.ESc.
> > Research and Development Engineer
> > Kitware Inc.,
> > North Carolina Office
> > www.kitware.com
> > 919-969-6990 X 319
> >
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110617/499e4c24/attachment.htm>

From haehn at bwh.harvard.edu  Fri Jun 17 17:05:27 2011
From: haehn at bwh.harvard.edu (Daniel Haehn)
Date: Fri, 17 Jun 2011 13:05:27 -0400
Subject: [Ctk-developers] Setting name and description of
	ctkWorkflowWidgetStep
In-Reply-To: <BANLkTi=sPbsoUkJ2tSAoU_uQ+aN2dygLPw@mail.gmail.com>
References: <BANLkTinEdZWWiCXvUhZ4TRMZOBnNyjHvTA@mail.gmail.com>
	<BANLkTin8iEVkh+fh0PiK-NV+0bbca+HNoQ@mail.gmail.com>
	<BANLkTimJ58=syYEvZqNc7YNT7G0xNZBg_A@mail.gmail.com>
	<BANLkTi==X0WQKtO636fLkRqaaMqiU752LA@mail.gmail.com>
	<BANLkTi=E1ZOs0FBuDNOJxMr+nHfK6re=KQ@mail.gmail.com>
	<BANLkTinMLEtQvjRcOuq-f9VaV6NvCnteMw@mail.gmail.com>
	<BANLkTi=sYr6megrCtJR3b2nE2Tj4mpsdTw@mail.gmail.com>
	<BANLkTi=sPbsoUkJ2tSAoU_uQ+aN2dygLPw@mail.gmail.com>
Message-ID: <BANLkTikA7-NXHRGWoY+ZKMJX-SXw6vqFMQ@mail.gmail.com>

Sure, will do.

Daniel

On Fri, Jun 17, 2011 at 12:25 PM, Jean-Christophe Fillion-Robin
<jchris.fillionr at kitware.com> wrote:
> Hi Daniel,
>
> Can we consider the case closed ?
>
> Could you document the issue on the bugtracker (label: Feature Request) ?
> See https://github.com/commontk/CTK/issues
>
> Thanks
> Jc
>
> On Sat, Jun 4, 2011 at 10:41 PM, Daniel Haehn <haehn at bwh.harvard.edu> wrote:
>>
>> Hi Danielle,
>>
>> that is absolutely correct!
>>
>> Maybe it would be easiest to just fire a
>> d->workflow->currentStepChanged(this) signal when
>> ctkWorkflowStep::setName or setDescription are called? Then, the
>> groupbox would be automatically updated.
>>
>> Nevertheless, it has no high priority for me since I now set the name
>> and description of the step in its constructor - before adding it to
>> the workflow.
>>
>> Cheers,
>> Daniel
>>
>> On Fri, Jun 3, 2011 at 4:13 PM, Danielle Pace <danielle.pace at kitware.com>
>> wrote:
>> > Hi Daniel,
>> > Actually, I think you exposed something important here.
>> > I'm assuming, from python, that you wrote something similar to:
>> > - create step 1 (with no name/description)
>> > - create step 2 (with no name/description)
>> > - create workflow widget and add steps
>> > - display workflow widget
>> > - set name/description for steps 1 and 2
>> > - expected to see workflow update, and it didn't
>> > Is this correct? ?I think we are missing some signals on ctkWorkflowStep
>> > such as NameChanged(QString) and DescriptionChanged(QString) that the
>> > ctkWorkflowGroupBox should be listening to.
>> > Perhaps you could still post your snippet of python code that wasn't
>> > working, so I can verify that it is an ordering of commands that we
>> > should
>> > support?
>> > Thanks,
>> >
>> > Danielle
>> >
>> >
>> >
>> > On Fri, Jun 3, 2011 at 3:57 PM, Daniel Haehn <haehn at bwh.harvard.edu>
>> > wrote:
>> >>
>> >> Hi Danielle,
>> >>
>> >> thanks to your response I checked it again and you are right: it was
>> >> my mistake :)
>> >>
>> >> If the name and description of the steps are set before the steps are
>> >> added to the workflow, it all works as expected.
>> >>
>> >> I set the stuff after adding it and of course it did not recognize it
>> >> until the signal was fired.
>> >>
>> >> Thanks for the great work!
>> >>
>> >> Cheers,
>> >> Daniel
>> >>
>> >> On Fri, Jun 3, 2011 at 3:44 PM, Danielle Pace
>> >> <danielle.pace at kitware.com>
>> >> wrote:
>> >> > Hi Daniel,
>> >> > You shouldn't have to call the currenStepChanged signal directly.
>> >> > ?Can
>> >> > you
>> >> > please post your bit of code that is not working to the CTK
>> >> > developers
>> >> > mailing list, and I will take a look?
>> >> > Thanks,
>> >> >
>> >> > Danielle
>> >> >
>> >> > On Fri, Jun 3, 2011 at 3:21 PM, Julien Finet
>> >> > <julien.finet at kitware.com>
>> >> > wrote:
>> >> >>
>> >> >>
>> >> >> ---------- Forwarded message ----------
>> >> >> From: Daniel Haehn <haehn at bwh.harvard.edu>
>> >> >> Date: Fri, Jun 3, 2011 at 3:13 PM
>> >> >> Subject: Re: [Ctk-developers] Setting name and description of
>> >> >> ctkWorkflowWidgetStep
>> >> >> To: ctk-developers at commontk.org
>> >> >>
>> >> >>
>> >> >> Hi guys,
>> >> >>
>> >> >> I got it:
>> >> >>
>> >> >> the ctkWorkflowGroupBox listens for the currentStepChanged signal to
>> >> >> update the title (==name) and subTitle (==description) for the
>> >> >> current
>> >> >> step.
>> >> >>
>> >> >> So maybe it makes sense to invoke this signal when the
>> >> >> ctkWorkflow.start() gets called?
>> >> >>
>> >> >> If I do it manually like that, it works:
>> >> >>
>> >> >> ? ? ?self.workflow.start()
>> >> >> ? ? ?self.workflow.currentStepChanged(steps[0])
>> >> >>
>> >> >> Cheers,
>> >> >> Daniel
>> >> >>
>> >> >> On Fri, Jun 3, 2011 at 12:17 PM, Daniel Haehn
>> >> >> <haehn at bwh.harvard.edu>
>> >> >> wrote:
>> >> >> > Hi devels,
>> >> >> >
>> >> >> > in C++ I can do
>> >> >> >
>> >> >> > ?this->setName("Title of step");
>> >> >> > ?this->setDescription("Instructions");
>> >> >> >
>> >> >> > in a ctkWorkflowWidgetStep, which then get nicely displayed in the
>> >> >> > associated ctkWorkflowGroupBox.
>> >> >> >
>> >> >> > In Python, the same methods can be accessed but the
>> >> >> > ctkWorkflowGroupBox does not seem to show them.
>> >> >> >
>> >> >> > Do I need to trigger an update on the box or something similar?
>> >> >> >
>> >> >> > Thank you!
>> >> >> > Daniel
>> >> >> >
>> >> >> _______________________________________________
>> >> >> Ctk-developers mailing list
>> >> >> Ctk-developers at commontk.org
>> >> >> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>> >> >>
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Danielle Pace, M.ESc.
>> >> > Research and Development Engineer
>> >> > Kitware Inc.,
>> >> > North Carolina Office
>> >> > www.kitware.com
>> >> > 919-969-6990 X 319
>> >> >
>> >
>> >
>> >
>> > --
>> > Danielle Pace, M.ESc.
>> > Research and Development Engineer
>> > Kitware Inc.,
>> > North Carolina Office
>> > www.kitware.com
>> > 919-969-6990 X 319
>> >
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
> --
> +1 919 869 8849
>
>


From s.zelzer at dkfz-heidelberg.de  Sat Jun 18 11:56:29 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Sat, 18 Jun 2011 13:56:29 +0200
Subject: [Ctk-developers] MinGW compatibility
Message-ID: <4DFC926D.5050109@dkfz-heidelberg.de>

Hi,

Just for your information, I back-ported some MinGW-related fixes in 
DCMTK 3.6.0 to the DCMTK fork in CTK. CTK can now be build with MinGW 
and DCMTK enabled.

Since back-porting is not really a nice job, I also want to emphasize 
that using the official DCMTK version in CTK is quite important.

Thanks,
Sascha


From dicom at offis.de  Mon Jun 20 07:49:17 2011
From: dicom at offis.de (OFFIS DICOM Team)
Date: Mon, 20 Jun 2011 09:49:17 +0200
Subject: [Ctk-developers] CTK and DCMTK 3.6 compatibility
In-Reply-To: <4DFB6918.6060304@dkfz.de>
References: <4DFB6918.6060304@dkfz.de>
Message-ID: <4DFEFB7D.2020702@offis.de>

Hi Marco and everybody,

sorry, I already sent on Friday but used my personal email adresss as a
sender and thus the mail was rejected:

Am 17.06.2011 16:47, schrieb Marco Nolden:
> Hi,
>
> I started using the CTK DICOM Widgets in MITK. Since we want to use the
> official 3.6 release of DCMTK there are some compatibility issues. To get
> started I fixed the constant names, adapted the superbuild and put this
> in a new branch
>
> https://github.com/commontk/CTK/commits/dcmtk-3.6

I tried the same a few weeks ago and I can confirm that only the constant
names need some work (like 10 minutes or so). This is because the DICOM
standard introduced official short names for attributes and we changed our
own system to the one in the standard (which was 99% the same, anyway).

I would recommend just switching not to 3.6. release but to our git
repository (I tried that in the superbuild and it seems to work fine), since
the code is pretty stable and we try to fix serious issues as soon as they
are reported.

> As far as I can see there are no build issues with DCMTK itself, the
> stock CMake files of DCMTK 3.6 seem to work fine. However the interface
> of the DcmSCU class has changed and I'm not sure how to replace the
> methods correctly and whether the implementation in 3.6 is already
> complete.

The implementation is already complete from the perspective of what you need
in the SCU, i.e. MOVE is implemted in particular. I can send you a patch
within the next days, Marco and Steve, that is about to adapt the current
CTK SCU stuff to the DcmSCU API (which I also already have on my
laptop but not tested yet).

Best regards,
Michael

-- 
OFFIS DICOM Team, Escherweg 2, 26121 Oldenburg, Germany
E-Mail: dicom at offis.de, URL: http://dicom.offis.de


From pieper at ibility.net  Mon Jun 20 15:02:41 2011
From: pieper at ibility.net (Steve Pieper)
Date: Mon, 20 Jun 2011 11:02:41 -0400
Subject: [Ctk-developers] CTK and DCMTK 3.6 compatibility
In-Reply-To: <4DFEFB7D.2020702@offis.de>
References: <4DFB6918.6060304@dkfz.de>
	<4DFEFB7D.2020702@offis.de>
Message-ID: <BANLkTi=m623c+wii8toGq3VOtSoDuJ_g2A@mail.gmail.com>

Hi Guys -

I also looked at the issues related to upgrading ctk's dcmtk to 3.6
and got stalled on adapting to the latest DcmSC(U/P) changes.  I'm
very much looking forward to Michael's patch :D.

In the mean time we've been focused working on the GUI side and
Nicholas has made great progress with the thumbnails and the image
viewer.  We'll be working on it this week at the na-mic project week
so hopefully we'll make some more useful progress (any suggestions
and/or feature requests welcome).

http://www.na-mic.org/Wiki/index.php/2011_Summer_Project_Week

http://www.na-mic.org/Wiki/index.php/2011_Summer_Project_Week_DICOM_Networking

Best regards,
Steve

On Mon, Jun 20, 2011 at 3:49 AM, OFFIS DICOM Team <dicom at offis.de> wrote:
> Hi Marco and everybody,
>
> sorry, I already sent on Friday but used my personal email adresss as a
> sender and thus the mail was rejected:
>
> Am 17.06.2011 16:47, schrieb Marco Nolden:
>>
>> Hi,
>>
>> I started using the CTK DICOM Widgets in MITK. Since we want to use the
>> official 3.6 release of DCMTK there are some compatibility issues. To get
>> started I fixed the constant names, adapted the superbuild and put this
>> in a new branch
>>
>> https://github.com/commontk/CTK/commits/dcmtk-3.6
>
> I tried the same a few weeks ago and I can confirm that only the constant
> names need some work (like 10 minutes or so). This is because the DICOM
> standard introduced official short names for attributes and we changed our
> own system to the one in the standard (which was 99% the same, anyway).
>
> I would recommend just switching not to 3.6. release but to our git
> repository (I tried that in the superbuild and it seems to work fine), since
> the code is pretty stable and we try to fix serious issues as soon as they
> are reported.
>
>> As far as I can see there are no build issues with DCMTK itself, the
>> stock CMake files of DCMTK 3.6 seem to work fine. However the interface
>> of the DcmSCU class has changed and I'm not sure how to replace the
>> methods correctly and whether the implementation in 3.6 is already
>> complete.
>
> The implementation is already complete from the perspective of what you need
> in the SCU, i.e. MOVE is implemted in particular. I can send you a patch
> within the next days, Marco and Steve, that is about to adapt the current
> CTK SCU stuff to the DcmSCU API (which I also already have on my
> laptop but not tested yet).
>
> Best regards,
> Michael
>
> --
> OFFIS DICOM Team, Escherweg 2, 26121 Oldenburg, Germany
> E-Mail: dicom at offis.de, URL: http://dicom.offis.de
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>


From julien.finet at kitware.com  Sun Jun 26 18:40:19 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Sun, 26 Jun 2011 14:40:19 -0400
Subject: [Ctk-developers] Filenames alphabetically sorted in CMakeLists.txt
Message-ID: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>

Hi CTK developers,

I'm not sure we ever discussed it before (I think I remember Marco talking
about it sometime) but I'd like to make sure we all agree that "files
inclusions must be alphabetically ordered in CMakeLists.txt" (except if
there is an obvious different category ordering).
This is helpful for 2 reasons:
 - It is less prone to generate conflicts when rebasing. If two commits add
a file, they are less unlikely to conflict if they are alphabetically
ordered than if they both are added at the bottom of the list.
 - (Faster to lookup a file in the cmakelists)

Of course, the discussion is open, so please share your thoughts !

Regards,
Julien.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110626/56009798/attachment.htm>

From jchris.fillionr at kitware.com  Sun Jun 26 20:14:34 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Sun, 26 Jun 2011 16:14:34 -0400
Subject: [Ctk-developers] Filenames alphabetically sorted in
	CMakeLists.txt
In-Reply-To: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>
References: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>
Message-ID: <BANLkTint8hcRzpf91KBOU_MNaGskKiujuA@mail.gmail.com>

+1

On Sun, Jun 26, 2011 at 2:40 PM, Julien Finet <julien.finet at kitware.com>wrote:

> Hi CTK developers,
>
> I'm not sure we ever discussed it before (I think I remember Marco talking
> about it sometime) but I'd like to make sure we all agree that "files
> inclusions must be alphabetically ordered in CMakeLists.txt" (except if
> there is an obvious different category ordering).
> This is helpful for 2 reasons:
>  - It is less prone to generate conflicts when rebasing. If two commits add
> a file, they are less unlikely to conflict if they are alphabetically
> ordered than if they both are added at the bottom of the list.
>  - (Faster to lookup a file in the cmakelists)
>
> Of course, the discussion is open, so please share your thoughts !
>
> Regards,
> Julien.
>
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>


-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110626/117d6fd2/attachment.htm>

From pieper at ibility.net  Mon Jun 27 14:57:01 2011
From: pieper at ibility.net (Steve Pieper)
Date: Mon, 27 Jun 2011 10:57:01 -0400
Subject: [Ctk-developers] Filenames alphabetically sorted in
	CMakeLists.txt
In-Reply-To: <BANLkTint8hcRzpf91KBOU_MNaGskKiujuA@mail.gmail.com>
References: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>
	<BANLkTint8hcRzpf91KBOU_MNaGskKiujuA@mail.gmail.com>
Message-ID: <BANLkTimg0iYrh4XAK6GTJk+GZ-C-OAscdQ@mail.gmail.com>

Sounds very good to me -- in fact, it looks like you added this to the
coding style back in February, Julien :D

http://www.commontk.org/index.php?title=Contributing_to_CTK&diff=669&oldid=668

On Sun, Jun 26, 2011 at 4:14 PM, Jean-Christophe Fillion-Robin <
jchris.fillionr at kitware.com> wrote:

> +1
>
> On Sun, Jun 26, 2011 at 2:40 PM, Julien Finet <julien.finet at kitware.com>wrote:
>
>> Hi CTK developers,
>>
>> I'm not sure we ever discussed it before (I think I remember Marco talking
>> about it sometime) but I'd like to make sure we all agree that "files
>> inclusions must be alphabetically ordered in CMakeLists.txt" (except if
>> there is an obvious different category ordering).
>> This is helpful for 2 reasons:
>>  - It is less prone to generate conflicts when rebasing. If two commits
>> add a file, they are less unlikely to conflict if they are alphabetically
>> ordered than if they both are added at the bottom of the list.
>>  - (Faster to lookup a file in the cmakelists)
>>
>> Of course, the discussion is open, so please share your thoughts !
>>
>> Regards,
>> Julien.
>>
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
>
>
> --
> +1 919 869 8849
>
>
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110627/c8e675c4/attachment.htm>

From julien.finet at kitware.com  Mon Jun 27 15:13:44 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Mon, 27 Jun 2011 11:13:44 -0400
Subject: [Ctk-developers] Filenames alphabetically sorted in
	CMakeLists.txt
In-Reply-To: <BANLkTimg0iYrh4XAK6GTJk+GZ-C-OAscdQ@mail.gmail.com>
References: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>
	<BANLkTint8hcRzpf91KBOU_MNaGskKiujuA@mail.gmail.com>
	<BANLkTimg0iYrh4XAK6GTJk+GZ-C-OAscdQ@mail.gmail.com>
Message-ID: <BANLkTi=HpRamXx0oFtLc18OUO7VeKmd-bA@mail.gmail.com>

I can't remember what I did in February, that's scary.... :-(

So my question becomes then a reminder to our CTK developers:
Please sort file names in CMakeLists ! :-P

I'll go ahead and fix the DICOM CMakeLists, it's not consistently sorted
there.

Thanks,
Julien.

On Mon, Jun 27, 2011 at 10:57 AM, Steve Pieper <pieper at ibility.net> wrote:

> Sounds very good to me -- in fact, it looks like you added this to the
> coding style back in February, Julien :D
>
>
> http://www.commontk.org/index.php?title=Contributing_to_CTK&diff=669&oldid=668
>
>
> On Sun, Jun 26, 2011 at 4:14 PM, Jean-Christophe Fillion-Robin <
> jchris.fillionr at kitware.com> wrote:
>
>> +1
>>
>> On Sun, Jun 26, 2011 at 2:40 PM, Julien Finet <julien.finet at kitware.com>wrote:
>>
>>> Hi CTK developers,
>>>
>>> I'm not sure we ever discussed it before (I think I remember Marco
>>> talking about it sometime) but I'd like to make sure we all agree that
>>> "files inclusions must be alphabetically ordered in CMakeLists.txt" (except
>>> if there is an obvious different category ordering).
>>> This is helpful for 2 reasons:
>>>  - It is less prone to generate conflicts when rebasing. If two commits
>>> add a file, they are less unlikely to conflict if they are alphabetically
>>> ordered than if they both are added at the bottom of the list.
>>>  - (Faster to lookup a file in the cmakelists)
>>>
>>> Of course, the discussion is open, so please share your thoughts !
>>>
>>> Regards,
>>> Julien.
>>>
>>> _______________________________________________
>>> Ctk-developers mailing list
>>> Ctk-developers at commontk.org
>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>
>>>
>>
>>
>> --
>> +1 919 869 8849
>>
>>
>> _______________________________________________
>> Ctk-developers mailing list
>> Ctk-developers at commontk.org
>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110627/ac7f969f/attachment.htm>

From jchris.fillionr at kitware.com  Mon Jun 27 15:25:00 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Mon, 27 Jun 2011 11:25:00 -0400
Subject: [Ctk-developers] Filenames alphabetically sorted in
	CMakeLists.txt
In-Reply-To: <BANLkTi=HpRamXx0oFtLc18OUO7VeKmd-bA@mail.gmail.com>
References: <BANLkTin9GzODvfBBU09wp16pB9DfR1zpgg@mail.gmail.com>
	<BANLkTint8hcRzpf91KBOU_MNaGskKiujuA@mail.gmail.com>
	<BANLkTimg0iYrh4XAK6GTJk+GZ-C-OAscdQ@mail.gmail.com>
	<BANLkTi=HpRamXx0oFtLc18OUO7VeKmd-bA@mail.gmail.com>
Message-ID: <BANLkTin7CySAaCAj-K=JtXidLwG8arbrEg@mail.gmail.com>

Note also that "#include" should also be sorted / organized by category.

For example:

// Qt includes
#include <QDebug>
#include <QPushButton>

// CTKAwesomeLib includes
...

// VTK includes
...

Jc

On Mon, Jun 27, 2011 at 11:13 AM, Julien Finet <julien.finet at kitware.com>wrote:

> I can't remember what I did in February, that's scary.... :-(
>
> So my question becomes then a reminder to our CTK developers:
> Please sort file names in CMakeLists ! :-P
>
> I'll go ahead and fix the DICOM CMakeLists, it's not consistently sorted
> there.
>
> Thanks,
> Julien.
>
>
> On Mon, Jun 27, 2011 at 10:57 AM, Steve Pieper <pieper at ibility.net> wrote:
>
>> Sounds very good to me -- in fact, it looks like you added this to the
>> coding style back in February, Julien :D
>>
>>
>> http://www.commontk.org/index.php?title=Contributing_to_CTK&diff=669&oldid=668
>>
>>
>> On Sun, Jun 26, 2011 at 4:14 PM, Jean-Christophe Fillion-Robin <
>> jchris.fillionr at kitware.com> wrote:
>>
>>> +1
>>>
>>> On Sun, Jun 26, 2011 at 2:40 PM, Julien Finet <julien.finet at kitware.com>wrote:
>>>
>>>> Hi CTK developers,
>>>>
>>>> I'm not sure we ever discussed it before (I think I remember Marco
>>>> talking about it sometime) but I'd like to make sure we all agree that
>>>> "files inclusions must be alphabetically ordered in CMakeLists.txt" (except
>>>> if there is an obvious different category ordering).
>>>> This is helpful for 2 reasons:
>>>>  - It is less prone to generate conflicts when rebasing. If two commits
>>>> add a file, they are less unlikely to conflict if they are alphabetically
>>>> ordered than if they both are added at the bottom of the list.
>>>>  - (Faster to lookup a file in the cmakelists)
>>>>
>>>> Of course, the discussion is open, so please share your thoughts !
>>>>
>>>> Regards,
>>>> Julien.
>>>>
>>>> _______________________________________________
>>>> Ctk-developers mailing list
>>>> Ctk-developers at commontk.org
>>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>>
>>>>
>>>
>>>
>>> --
>>> +1 919 869 8849
>>>
>>>
>>> _______________________________________________
>>> Ctk-developers mailing list
>>> Ctk-developers at commontk.org
>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>
>>>
>>
>


-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110627/8c2c0d6e/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Thu Jun 30 15:19:11 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Thu, 30 Jun 2011 17:19:11 +0200
Subject: [Ctk-developers] MacOS 10.6 errors
Message-ID: <4E0C93EF.2060303@dkfz-heidelberg.de>

Hi guys,

We are having errors under MacOS 10.6 because the file 
CMake/ctkBlockSetCMakeOSXVariables.cmake hard-codes a reference to a 
MacOSX 10.5 SDK.

Any ideas?

Thanks,
Sascha


From jchris.fillionr at kitware.com  Thu Jun 30 15:40:19 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Thu, 30 Jun 2011 11:40:19 -0400
Subject: [Ctk-developers] MacOS 10.6 errors
In-Reply-To: <4E0C93EF.2060303@dkfz-heidelberg.de>
References: <4E0C93EF.2060303@dkfz-heidelberg.de>
Message-ID: <BANLkTikExskvD+eWA1mhM_bwCk4c_t33MA@mail.gmail.com>

Just backported proper fix from Slicer.
See
https://github.com/commontk/CTK/commit/cc9fa878086128e9cfd70f007ffcde6592470638

Let me know if you experience any issue.
Thanks
Jc

On Thu, Jun 30, 2011 at 11:19 AM, Sascha Zelzer <s.zelzer at dkfz-heidelberg.de
> wrote:

> Hi guys,
>
> We are having errors under MacOS 10.6 because the file CMake/**
> ctkBlockSetCMakeOSXVariables.**cmake hard-codes a reference to a MacOSX
> 10.5 SDK.
>
> Any ideas?
>
> Thanks,
> Sascha
> ______________________________**_________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-**bin/mailman/listinfo/ctk-**developers<http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers>
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110630/42652d85/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Thu Jun 30 17:07:57 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Thu, 30 Jun 2011 19:07:57 +0200
Subject: [Ctk-developers] MacOS 10.6 errors
In-Reply-To: <BANLkTikExskvD+eWA1mhM_bwCk4c_t33MA@mail.gmail.com>
References: <4E0C93EF.2060303@dkfz-heidelberg.de>
	<BANLkTikExskvD+eWA1mhM_bwCk4c_t33MA@mail.gmail.com>
Message-ID: <4E0CAD6D.4040007@dkfz-heidelberg.de>

Hi Jc,

That was fast. Works perfect on our MacOS installation!

Thanks a lot,

Sascha

On 06/30/2011 05:40 PM, Jean-Christophe Fillion-Robin wrote:
> Just backported proper fix from Slicer.
> See 
> https://github.com/commontk/CTK/commit/cc9fa878086128e9cfd70f007ffcde6592470638
>
> Let me know if you experience any issue.
> Thanks
> Jc
>
> On Thu, Jun 30, 2011 at 11:19 AM, Sascha Zelzer 
> <s.zelzer at dkfz-heidelberg.de <mailto:s.zelzer at dkfz-heidelberg.de>> wrote:
>
>     Hi guys,
>
>     We are having errors under MacOS 10.6 because the file
>     CMake/ctkBlockSetCMakeOSXVariables.cmake hard-codes a reference to
>     a MacOSX 10.5 SDK.
>
>     Any ideas?
>
>     Thanks,
>     Sascha
>     _______________________________________________
>     Ctk-developers mailing list
>     Ctk-developers at commontk.org <mailto:Ctk-developers at commontk.org>
>     http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>
>
>
>
> -- 
> +1 919 869 8849
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20110630/1afdc13b/attachment.htm>