From espakm at gmail.com  Tue Nov  1 16:58:03 2011
From: espakm at gmail.com (Miklos Espak)
Date: Tue, 1 Nov 2011 17:58:03 +0100
Subject: [Ctk-developers] workflow framework and threads
In-Reply-To: <CAOVq=oP2bbefVaWq+F7XzZHPi-4FYhVaZEbWDipaY4u7PJgZig@mail.gmail.com>
References: <CAOVq=oPFHUZ5vsz7_5LcwUeMSnmBTL9GfNV-RptU=qLba1+shw@mail.gmail.com>
	<CANLz8R4sGkqSOAkYz31AM7eEzG9j9dzB96pr_6p1VSDn-5gyEw@mail.gmail.com>
	<CAOVq=oNokJeTOEZnJcMW_gyVZmCR8kVae3FeKWF2LCfi7bfYmQ@mail.gmail.com>
	<CAOVq=oP2bbefVaWq+F7XzZHPi-4FYhVaZEbWDipaY4u7PJgZig@mail.gmail.com>
Message-ID: <CAOVq=oMf5GexOkMNjPLFzK61us28Rmw4BoCszB9nE-=v42KFYw@mail.gmail.com>

Hi,

could you (or anybody), eventually, have a look at the test that I shared?

As a workaround I disabled the Next button until the computation finishes
(between the onExit and the onExitComplete signal), but it treats only one
symptom.

Now I ran into the problem again, and I do not know what to do with this.
Step2 relies on some data that Step1 creates in the onExitComplete
function. However, because of the bug, the data is not ready. The events
happen in this order:

Step1 onExit
Step2 onEntry
Step2 createUserInterface
Step1 onExitComplete

In my opinion, the step1 onExitComplete should be invoked before step2
onEntry.

Even, it would not be bad if the user interface was ready before the
onEntry signal, but it is not so important.

Thanks,
Miklos

On Tue, Oct 25, 2011 at 3:11 PM, Miklos Espak <espakm at gmail.com> wrote:

> Hi,
>
> I added a new test. You can run it by
>
> $ ./CTKWidgetsCppTests ctkExampleUseOfWorkflowWidgetUsingThreads -I
>
> in the project bin directory.
>
> The onExit function of the steps start a thread that waits for 5 sec
> ("calculation"). The onExitComplete signal is sent when the thread finishes.
>
> If you press the Next button, the "calculation" starts and after 5 sec you
> get to the next page. That is fine. However, if you press Next within this
> period, the thread of the second step will start, although the first step
> has not finished. (It has not sent the onExitComplete signal.)
>
> You can checkout the test via GitHub:
>
> https://github.com/commontk/CTK/issues/61
>
> Thanks for dealing with this issue!
>
> Best,
> Miklos
>
> On Mon, Oct 24, 2011 at 11:47 PM, Miklos Espak <espakm at gmail.com> wrote:
>
>> By running the step in a different thread, what do you mean exactly ?
>>>
>>
>> Sorry, I was not clear. I run some calculations (ITK filters) in the
>> workflow steps, and I start them in a new thread. The workflow step listens
>> to the "finished" and "terminated" signal of the thread and it signals
>> "onExitCompleted" after the thread is finished.
>>
>>
>>> As a side note: "Although QObject<http://doc.qt.nokia.com/stable/qobject.html>is reentrant, the GUI classes, notably
>>> QWidget <http://doc.qt.nokia.com/stable/qwidget.html> and all its
>>> subclasses, are not reentrant. They can only be used from the main thread."
>>> See http://doc.qt.nokia.com/stable/threads-qobject.html
>>>
>>
>> Yes, I had some troubles because of this, but I solved them. I refactored
>> every GUI access from the function that runs the calculation. E.g. the
>> thread that runs the calculation notifies the GUI thread through signals
>> about the progress of the internal ITK filters. In this way the thread does
>> not need to access the progress bar.
>>
>> Would it be possible for you to share some code ? May a topic on your CTK
>>>> fork ?
>>>>
>>>
>> Yes, I will upload a sample app tomorrow.
>>
>> Thanks,
>> Miklos
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111101/e5fde796/attachment.htm>

From dicom at offis.de  Wed Nov  2 15:33:47 2011
From: dicom at offis.de (OFFIS DICOM Team)
Date: Wed, 02 Nov 2011 16:33:47 +0100
Subject: [Ctk-developers] how to enable logging
In-Reply-To: <CAOVq=oO5Mznkc+-J+BihOd9iqTVyuZESE81cHLtiOwy3mNN-7Q@mail.gmail.com>
References: <CAOVq=oO5Mznkc+-J+BihOd9iqTVyuZESE81cHLtiOwy3mNN-7Q@mail.gmail.com>
Message-ID: <4EB162DB.9010802@offis.de>

Hi Miklos,

just returned from vacation, so this answer is pretty late:

Am 21.10.2011 19:06, schrieb Miklos Espak:
> Hi,
>
> how to enable logging in CTK? I enabled the Log4Qt option in CTK and
> rebuilt it, but it is apparently not enough. Do I have to create a config
> file? The syntax is like for Log4J?

If it is relevant for you to also configure the DICOM-related logging the
logging detail, i.e. format, destination and so on, you might be interested
in little howtos in our wiki:

   http://support.dcmtk.org/wiki/dcmtk/howto/start#logging

All the best,
Michael

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


From espakm at gmail.com  Wed Nov  2 15:47:56 2011
From: espakm at gmail.com (Miklos Espak)
Date: Wed, 2 Nov 2011 16:47:56 +0100
Subject: [Ctk-developers] how to enable logging
In-Reply-To: <4EB162DB.9010802@offis.de>
References: <CAOVq=oO5Mznkc+-J+BihOd9iqTVyuZESE81cHLtiOwy3mNN-7Q@mail.gmail.com>
	<4EB162DB.9010802@offis.de>
Message-ID: <CAOVq=oPhBcM5Y0uwiaBO7q3=wWMyZHvQyNBNV0owMk-uR0f8wg@mail.gmail.com>

Hi Michael,

Jean-Christophe gave the solution, it works now.
Thanks for the link, I did not know that page, it contains many useful
howtos.

Thanks,
Miklos
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111102/7a532a91/attachment.htm>

From pieper at ibility.net  Wed Nov  9 20:58:46 2011
From: pieper at ibility.net (Steve Pieper)
Date: Wed, 9 Nov 2011 15:58:46 -0500
Subject: [Ctk-developers] Fwd: [Na-mic-project-week] Invitation to 2011
 Winter Project Week in Salt Lake City
In-Reply-To: <4EA8AF7E.3070604@bwh.harvard.edu>
References: <4EA8AF7E.3070604@bwh.harvard.edu>
Message-ID: <CA+OFd4ydKfEM_L4wD_X-SBL-FpWU+p5z+Q_syiaBoOp6QqEdcw@mail.gmail.com>

Hi CTKers -

For people interested in more hacking opportunities we wanted to send
around an announcement/reminder of the upcoming na-mic meeting in Salt Lake
City, Utah in early January.  This will be the 14th time we've held this
event, and it never fails to be a stimulating and rewarding time to meet
with like-minded developers and work on thorny projects that can best be
solved with some face to face time.

To get an idea of the atmosphere and the types of projects, have a look at
this page about the most recent summer's event:

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

For logistics and other info about the upcoming event look here:

  http://www.na-mic.org/Wiki/index.php/AHM_2012

Thanks,
Steve


---------- Forwarded message ----------
From: Tina Kapur <tkapur at bwh.harvard.edu>
Date: Wed, Oct 26, 2011 at 9:10 PM
Subject: [Na-mic-project-week] Invitation to 2011 Winter Project Week in
Salt Lake City
To: na-mic-project-week at na-mic.org, namic-all at na-mic.org,
igtws at bwh.harvard.edu


Dear Colleague,

We would like to invite you to the 14th Project Week in Salt Lake City,
Utah, from January 9-13, 2012.  Similar to previous events, we will spend
this week pursuing hands-on research and development activity in
Neuroscience and Image-Guided Therapy applications.

If you are interested in learning more about this the 14th Project Week
of hands-on research and software development, please join us for our
kick-off teleconference tomorrow i.e. Thursday, October 27th at 3pm ET.

Dial 800-728-9607 (or 913-904-9873) Participant code: 6815251#

Best Regards,
-Tina

-- 
Tina Kapur, Ph.D.
Executive Director
Image Guided Therapy Program
Department of Radiology
Brigham and Women's Hospital
75 Francis Street, Boston MA 02115





______________________________**_________________
Na-mic-project-week mailing list
Na-mic-project-week at na-mic.org
http://public.kitware.com/cgi-**bin/mailman/listinfo/na-mic-**project-week<http://public.kitware.com/cgi-bin/mailman/listinfo/na-mic-project-week>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111109/b9261627/attachment.htm>

From jchris.fillionr at kitware.com  Mon Nov 14 10:36:14 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Mon, 14 Nov 2011 05:36:14 -0500
Subject: [Ctk-developers] CTK - What about requiring CMake 2.8.4 ?
Message-ID: <CANLz8R6Q0FdhMvSk0O=-cUDqBvJid-Hk+WaQzPpszYPVELWgWQ@mail.gmail.com>

Folks,

Does it sound reasonable if CTK requires at least CMake 2.8.4 ?

The motivation behind that change is explained here:
https://github.com/commontk/CTK/issues/13

An alternative solution would be to provide a custom ExternalProject.cmake
within CTK repository. However, it would be great if we could avoid to
maintain our own version of ExternalProject.cmake.

Thanks for your input
Jc

-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111114/c63fc194/attachment.htm>

From jchris.fillionr at kitware.com  Mon Nov 14 13:24:17 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Mon, 14 Nov 2011 08:24:17 -0500
Subject: [Ctk-developers] CTK Hackfest - Google hangout
Message-ID: <CANLz8R6g8GuVwCjoKX+7XNopYvnkuCa7WaGPJYai9aQctqn5VA@mail.gmail.com>

A common place to share your screen, chat, etc ...
See https://plus.google.com/hangouts/extras/talk.google.com/CTK#

I already joined ..
Jc

-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111114/323f779e/attachment.htm>

From jchris.fillionr at kitware.com  Mon Nov 14 14:14:47 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Mon, 14 Nov 2011 09:14:47 -0500
Subject: [Ctk-developers] CTK - What about requiring CMake 2.8.4 ?
In-Reply-To: <CANLz8R6Q0FdhMvSk0O=-cUDqBvJid-Hk+WaQzPpszYPVELWgWQ@mail.gmail.com>
References: <CANLz8R6Q0FdhMvSk0O=-cUDqBvJid-Hk+WaQzPpszYPVELWgWQ@mail.gmail.com>
Message-ID: <CANLz8R5PKBOjnGTdSAYSnkw4pjATW=QPPxRU_Lyc59gKBTuOSg@mail.gmail.com>

Folks,

So far, the following project have no objection requiring CMake 2.8.4 for
CTK:
  - MedInria / DTK
  - Slicer
  - MITK
  - Maf3
  - GMIAS (Does not depend on CTK)

If there are no objections, I will change the required CMake version and
work on fixing the following issue:
https://github.com/commontk/CTK/issues/13

Thanks
Jc

On Mon, Nov 14, 2011 at 5:36 AM, Jean-Christophe Fillion-Robin <
jchris.fillionr at kitware.com> wrote:

> Folks,
>
> Does it sound reasonable if CTK requires at least CMake 2.8.4 ?
>
> The motivation behind that change is explained here:
> https://github.com/commontk/CTK/issues/13
>
> An alternative solution would be to provide a custom ExternalProject.cmake
> within CTK repository. However, it would be great if we could avoid to
> maintain our own version of ExternalProject.cmake.
>
> Thanks for your input
> Jc
>
> --
> +1 919 869 8849
>
>


-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111114/7c7f21b7/attachment.htm>

From pieper at ibility.net  Thu Nov 17 16:28:16 2011
From: pieper at ibility.net (Steve Pieper)
Date: Thu, 17 Nov 2011 11:28:16 -0500
Subject: [Ctk-developers] roadmap discussion
Message-ID: <CA+OFd4wt5e1Uh5osXHcoHLidC=90Q6ugpVFSoueoTrJvigOZsw@mail.gmail.com>

Hello from the Sofia-Antipolis Hackfest -

Between hacking sessions we found time to discuss and write up the roadmap
for CTK to lay out some of the key concepts, technologies, and goals of the
effort.  We hope this will help new users and contributors get a quick idea
of the CTK.  Further ideas and suggestions from the people who couldn't
make it to France is also welcome.   The roadmap is now linked from the
commontk.org home page.

http://www.commontk.org/index.php/CTK-Roadmap

-Steve

p.s. we're also writing a lot of code!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111117/d9e653c2/attachment.htm>

From s.zelzer at dkfz-heidelberg.de  Fri Nov 18 15:42:55 2011
From: s.zelzer at dkfz-heidelberg.de (Sascha Zelzer)
Date: Fri, 18 Nov 2011 16:42:55 +0100
Subject: [Ctk-developers] CMake configuration simplified
Message-ID: <4EC67CFF.2020809@dkfz-heidelberg.de>

Hi Folks,

I would like to draw your attention to a new feature of the CTK build 
system which I personally find very pleasing.

During the Sophia-Antipolis Hackfest we worked on making it easier for 
new CTK users to configure CTK with CMake. Previously, many build 
options were displayed after the initial configure run. This was 
overwhelming and confusing.

The new system retains all previous CMake options but marks nearly all 
of them as advanced. You will now see a set of high-level options which 
are used to enable a default set of libraries and/or plugins related to 
one of the main "sub-projects" inside CTK. However, you still have full 
control over all build options in advanced mode.

The relevant merge commit is

https://github.com/commontk/CTK/commit/7cac6550c2cade1fb69ec2136c89b56ed332d99b

Let us now if you have any troubles using it.

Thanks,

Sascha


From espakm at gmail.com  Wed Nov 23 07:35:07 2011
From: espakm at gmail.com (Miklos Espak)
Date: Wed, 23 Nov 2011 08:35:07 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
Message-ID: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>

Hi,

I created a few widget steps. I noticed that the user interface is
created lazily, at the first time it is needed, after the first
onEntry call.

For this reason I have to check in the onEntry function if the user
interface has been created, so I can update the GUI.

I have not found a nice way to do this.

There is a "createUserInterfaceComplete" function that "signals" if
the UI has been created. This is not a Qt signal, however, so you
cannot connect it to a slot.

It is not virtual either, so you cannot override it safely.

The only reasonable way I found was to create a
createUserInterfaceComplete2 function in the common superclass of my
widget steps, that calls createUserInterfaceComplete *and* sets a
flag.

Would it be possible to make one of these changes?
  - introduce an isUserInterfaceComplete() function or such
  - make createUserInterfaceComplete virtual
  - make createUserInterfaceComplete a Qt signal, so one can be
notified if the UI is created
  - create the GUI before the first onEntry call

Thanks,
Miklos


From julien.finet at kitware.com  Fri Nov 25 14:17:58 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Fri, 25 Nov 2011 15:17:58 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
Message-ID: <CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>

Hi Miklos,

Can you suggest a patch ?
We can review and merge it.

Regards,
Julien.

On Wed, Nov 23, 2011 at 8:35 AM, Miklos Espak <espakm at gmail.com> wrote:
> Hi,
>
> I created a few widget steps. I noticed that the user interface is
> created lazily, at the first time it is needed, after the first
> onEntry call.
>
> For this reason I have to check in the onEntry function if the user
> interface has been created, so I can update the GUI.
>
> I have not found a nice way to do this.
>
> There is a "createUserInterfaceComplete" function that "signals" if
> the UI has been created. This is not a Qt signal, however, so you
> cannot connect it to a slot.
>
> It is not virtual either, so you cannot override it safely.
>
> The only reasonable way I found was to create a
> createUserInterfaceComplete2 function in the common superclass of my
> widget steps, that calls createUserInterfaceComplete *and* sets a
> flag.
>
> Would it be possible to make one of these changes?
> ?- introduce an isUserInterfaceComplete() function or such
> ?- make createUserInterfaceComplete virtual
> ?- make createUserInterfaceComplete a Qt signal, so one can be
> notified if the UI is created
> ?- create the GUI before the first onEntry call
>
> Thanks,
> Miklos
> _______________________________________________
> Ctk-developers mailing list
> Ctk-developers at commontk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>


From espakm at gmail.com  Fri Nov 25 14:45:08 2011
From: espakm at gmail.com (Miklos Espak)
Date: Fri, 25 Nov 2011 15:45:08 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
	<CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
Message-ID: <CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>

Yes, I can create a patch.

Which solution you prefer? I would go for the first.

>> Would it be possible to make one of these changes?
>> ?- introduce an isUserInterfaceComplete() function or such
>> ?- make createUserInterfaceComplete virtual
>> ?- make createUserInterfaceComplete a Qt signal, so one can be
>> notified if the UI is created
>> ?- create the GUI before the first onEntry call
>>
>> Thanks,
>> Miklos
>> _______________________________________________
>> 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  Fri Nov 25 14:58:40 2011
From: julien.finet at kitware.com (Julien Finet)
Date: Fri, 25 Nov 2011 15:58:40 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
	<CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
	<CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>
Message-ID: <CAKoWPPF-s601g5emDeFEEJH-TDy6adtWPwXqt4sAidjf3uZt9A@mail.gmail.com>

Hi Miklos,

Looking at the code, it seems that the signal is fired:

//-----------------------------------------------------------------------------
void ctkWorkflowWidgetStepPrivate::createUserInterfaceCompleteInternal()const
{
  emit createUserInterfaceComplete();
}

It is done in the Pimpl though. If you search for
"createUserInterfaceComplete" in CTK, you'll find some use case.
To be honest, I've no idea how it works :-)
Maybe Jc or Danielle can comment on it.

I am surprised you need to listen for createUserInterfaceComplete, I
would expect to do all the initialization work in
ctkWorkflowWidgetStep::createUserInterface().

j.

On Fri, Nov 25, 2011 at 3:45 PM, Miklos Espak <espakm at gmail.com> wrote:
> Yes, I can create a patch.
>
> Which solution you prefer? I would go for the first.
>
>>> Would it be possible to make one of these changes?
>>> ?- introduce an isUserInterfaceComplete() function or such
>>> ?- make createUserInterfaceComplete virtual
>>> ?- make createUserInterfaceComplete a Qt signal, so one can be
>>> notified if the UI is created
>>> ?- create the GUI before the first onEntry call
>>>
>>> Thanks,
>>> Miklos
>>> _______________________________________________
>>> Ctk-developers mailing list
>>> Ctk-developers at commontk.org
>>> http://public.kitware.com/cgi-bin/mailman/listinfo/ctk-developers
>>>
>>
>


From espakm at gmail.com  Fri Nov 25 15:52:40 2011
From: espakm at gmail.com (Miklos Espak)
Date: Fri, 25 Nov 2011 16:52:40 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CAKoWPPF-s601g5emDeFEEJH-TDy6adtWPwXqt4sAidjf3uZt9A@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
	<CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
	<CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>
	<CAKoWPPF-s601g5emDeFEEJH-TDy6adtWPwXqt4sAidjf3uZt9A@mail.gmail.com>
Message-ID: <CAOVq=oO1pJjm2bf6JRCFsyz5EKpz1CUa9jvTr1t3zKR4k4SrHA@mail.gmail.com>

On Fri, Nov 25, 2011 at 3:58 PM, Julien Finet <julien.finet at kitware.com> wrote:
> Hi Miklos,
>
> Looking at the code, it seems that the signal is fired:
>
> //-----------------------------------------------------------------------------
> void ctkWorkflowWidgetStepPrivate::createUserInterfaceCompleteInternal()const
> {
> ?emit createUserInterfaceComplete();
> }
>
> It is done in the Pimpl though. If you search for
> "createUserInterfaceComplete" in CTK, you'll find some use case.
> To be honest, I've no idea how it works :-)
> Maybe Jc or Danielle can comment on it.
>
> I am surprised you need to listen for createUserInterfaceComplete, I
> would expect to do all the initialization work in
> ctkWorkflowWidgetStep::createUserInterface().

I have to fill the input widgets with values when I arrive at a
workflow step. I would do it in the onEntry, but at the first onEntry
call the user interface is not ready.

So I have to check in the onEntry if the GUI is created or not. This
cannot be queried by the ctkWorkflowStepWidget functions.

But no problem, I rearrange my code so I can decide it by a null check.

Best,
Miklos


From jchris.fillionr at kitware.com  Wed Nov 30 22:05:30 2011
From: jchris.fillionr at kitware.com (Jean-Christophe Fillion-Robin)
Date: Wed, 30 Nov 2011 17:05:30 -0500
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CAOVq=oO1pJjm2bf6JRCFsyz5EKpz1CUa9jvTr1t3zKR4k4SrHA@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
	<CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
	<CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>
	<CAKoWPPF-s601g5emDeFEEJH-TDy6adtWPwXqt4sAidjf3uZt9A@mail.gmail.com>
	<CAOVq=oO1pJjm2bf6JRCFsyz5EKpz1CUa9jvTr1t3zKR4k4SrHA@mail.gmail.com>
Message-ID: <CANLz8R7JK63qELeSoa94ng0haFs2O_G0bqMWqwDUpxRSX_Ovrg@mail.gmail.com>

Hi Miklos,

Keep us posted on your progress.

Thanks
Jc

On Fri, Nov 25, 2011 at 10:52 AM, Miklos Espak <espakm at gmail.com> wrote:

> On Fri, Nov 25, 2011 at 3:58 PM, Julien Finet <julien.finet at kitware.com>
> wrote:
> > Hi Miklos,
> >
> > Looking at the code, it seems that the signal is fired:
> >
> >
> //-----------------------------------------------------------------------------
> > void
> ctkWorkflowWidgetStepPrivate::createUserInterfaceCompleteInternal()const
> > {
> >  emit createUserInterfaceComplete();
> > }
> >
> > It is done in the Pimpl though. If you search for
> > "createUserInterfaceComplete" in CTK, you'll find some use case.
> > To be honest, I've no idea how it works :-)
> > Maybe Jc or Danielle can comment on it.
> >
> > I am surprised you need to listen for createUserInterfaceComplete, I
> > would expect to do all the initialization work in
> > ctkWorkflowWidgetStep::createUserInterface().
>
> I have to fill the input widgets with values when I arrive at a
> workflow step. I would do it in the onEntry, but at the first onEntry
> call the user interface is not ready.
>
> So I have to check in the onEntry if the GUI is created or not. This
> cannot be queried by the ctkWorkflowStepWidget functions.
>
> But no problem, I rearrange my code so I can decide it by a null check.
>
> Best,
> Miklos
>



-- 
+1 919 869 8849
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/ctk-developers/attachments/20111130/d119da1d/attachment.htm>

From espakm at gmail.com  Wed Nov 30 23:20:31 2011
From: espakm at gmail.com (Miklos Espak)
Date: Thu, 1 Dec 2011 00:20:31 +0100
Subject: [Ctk-developers] ctkWorkflowWidgetStep API
In-Reply-To: <CANLz8R7JK63qELeSoa94ng0haFs2O_G0bqMWqwDUpxRSX_Ovrg@mail.gmail.com>
References: <CAOVq=oPjQo9MvsoPshRKiXPA2SpskOjQNN7j=rmtar9Zh0oNZA@mail.gmail.com>
	<CAKoWPPGtnmDDy05uozqHkm75r7E3C_JAsSLKSwC24DmJ4a9N4Q@mail.gmail.com>
	<CAOVq=oMyfC=hQR8B62_dkZJ-nP-q7TYtb1wY_DA15KWBvxCAFA@mail.gmail.com>
	<CAKoWPPF-s601g5emDeFEEJH-TDy6adtWPwXqt4sAidjf3uZt9A@mail.gmail.com>
	<CAOVq=oO1pJjm2bf6JRCFsyz5EKpz1CUa9jvTr1t3zKR4k4SrHA@mail.gmail.com>
	<CANLz8R7JK63qELeSoa94ng0haFs2O_G0bqMWqwDUpxRSX_Ovrg@mail.gmail.com>
Message-ID: <CAOVq=oOHLzRt+J5ijZHuCyVM0TGC26briyKBX_NJiNatj+4dpQ@mail.gmail.com>

I found a trivial solution, by arranging my code. I keep a pointer to
the widget (called "ui") that I initialize it with 0 in the
constructor, and assign the real widget instance to it in the
createUserInterface function, so I can decide by a simple null pointer
check whether the user interface has been created. (Until now "ui" was
an ordinary variable, not a pointer, and I called ui.setupUi(this) in
the createUserInterface.)

Sorry for this many mails about this.

However, my other problem is still there. It was that sometimes the
onEntry function is not called at all. I uploaded a test to
reconstruct the situation. Have you checked it eventually?

It is here:
https://github.com/commontk/CTK/issues/61

Thanks,
Miklos

On Wed, Nov 30, 2011 at 11:05 PM, Jean-Christophe Fillion-Robin
<jchris.fillionr at kitware.com> wrote:
> Hi Miklos,
>
> Keep us posted on your progress.
>
> Thanks
> Jc
>
> On Fri, Nov 25, 2011 at 10:52 AM, Miklos Espak <espakm at gmail.com> wrote:
>>
>> On Fri, Nov 25, 2011 at 3:58 PM, Julien Finet <julien.finet at kitware.com>
>> wrote:
>> > Hi Miklos,
>> >
>> > Looking at the code, it seems that the signal is fired:
>> >
>> >
>> > //-----------------------------------------------------------------------------
>> > void
>> > ctkWorkflowWidgetStepPrivate::createUserInterfaceCompleteInternal()const
>> > {
>> > ?emit createUserInterfaceComplete();
>> > }
>> >
>> > It is done in the Pimpl though. If you search for
>> > "createUserInterfaceComplete" in CTK, you'll find some use case.
>> > To be honest, I've no idea how it works :-)
>> > Maybe Jc or Danielle can comment on it.
>> >
>> > I am surprised you need to listen for createUserInterfaceComplete, I
>> > would expect to do all the initialization work in
>> > ctkWorkflowWidgetStep::createUserInterface().
>>
>> I have to fill the input widgets with values when I arrive at a
>> workflow step. I would do it in the onEntry, but at the first onEntry
>> call the user interface is not ready.
>>
>> So I have to check in the onEntry if the GUI is created or not. This
>> cannot be queried by the ctkWorkflowStepWidget functions.
>>
>> But no problem, I rearrange my code so I can decide it by a null check.
>>
>> Best,
>> Miklos
>
>
>
> --
> +1 919 869 8849
>
>