[ITK-dev] [ITK] ITK API Change (was ITKApps failing after recent commit)
Bill Lorensen
bill.lorensen at gmail.com
Tue May 26 12:01:39 EDT 2015
Brad,
In an application, by separating the declaration (in a .h file) from
the definition (in a .cxx file) solves the problem. But this is a PITA
for an application that currently uses itkEventMacro. Upsetting
customers is not moving forward.
On Tue, May 26, 2015 at 11:58 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
> Brad,
>
> That is not the problem. The old macro duplicated implementations in
> each file that included a common .h file. The new api changes that.
> But, is it important. I think not. The old, less elegant form, worked.
>
>
> On Tue, May 26, 2015 at 11:44 AM, Bradley Lowekamp
> <blowekamp at mail.nih.gov> wrote:
>>
>> Lets try to keep moving forward.
>>
>> The itkEventMacro is not really used in ITK. It should not be hard to revert it to the old behavior.
>>
>> All that is likely needed is to improve the usage of the ITKEvent_EXPORT define [1]. It looks like the re-write of these macros was only intended for internal ITK usage as it's defined an ITKCommom_EXPORT.
>>
>> Brad
>>
>>
>>
>> [1] https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Core/Common/include/itkEventObject.h#L108
>>
>> On May 26, 2015, at 11:37 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>
>>> I agree. I will soon have a patch to the ITKApps that works for both
>>> versions of ITK. This will illustrate how much work is required to
>>> support both API's. The current ITK makes itkEventMacro dangerous and
>>> worthless.
>>>
>>> On Tue, May 26, 2015 at 11:31 AM, Matt McCormick
>>> <matt.mccormick at kitware.com> wrote:
>>>> Hi Bill,
>>>>
>>>> Thanks for bringing attention to this. If we cannot find a way for
>>>> external Event code to work without modification, I thing we should
>>>> revert the change before we release 4.8.
>>>>
>>>> 2 cents,
>>>> Matt
>>>>
>>>> On Tue, May 26, 2015 at 11:26 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>>>> itkEventMacro no longer works the way it used to. Applications that used
>>>>> itkEventMacro(myEvent, itk::UserEvent);
>>>>> must break up the statement into two statements and two files:
>>>>> in the .h file
>>>>> itkEventMacroDeclaration(myEvent,itk::UserEvent);
>>>>> in the .cxx file
>>>>> itkEventMacroDefinition(myEvent,itk::UserEvent);
>>>>>
>>>>> The problems occurs when an application has multiple source files that
>>>>> include the file with itkEventMacro. Duplicate symbols are created
>>>>> now. Prior to the API-break this was fine.
>>>>>
>>>>> Application code that uses itkEventMacro that wants to support ITK
>>>>> pre-the-API-break and post-the-API-break will need further statements.
>>>>>
>>>>> I am working on getting the ITKApps to work with both pre and post
>>>>> API-break versions.
>>>>>
>>>>> The unintended consequences of this API break are troublesome.
>>>>>
>>>>> On Mon, May 25, 2015 at 3:10 PM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>>>>> itksnap also has issues with the API change. I'll try to make a small
>>>>>> example that illustrates the problem.
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Mon, May 25, 2015 at 12:08 PM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>>>>>> Looks to me that the API for itkEventMacro has changed. This is bad.
>>>>>>> In ITKApps there are several definitions like this. How must these be
>>>>>>> changed?
>>>>>>>
>>>>>>> namespace fltk {
>>>>>>>
>>>>>>> /**
>>>>>>> * List of the events expected in this domain
>>>>>>> */
>>>>>>> itkEventMacro( RedrawEvent, itk::UserEvent );
>>>>>>> itkEventMacro( GlDrawEvent, itk::UserEvent );
>>>>>>> itkEventMacro( VolumeReslicedEvent, itk::UserEvent );
>>>>>>> itkEventMacro( ClippingPlaneEvent, itk::UserEvent );
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> } // end namespace fltk
>>>>>>>
>>>>>>>
>>>>>>> On Mon, May 25, 2015 at 11:59 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>>>>>>> Since this topic was merged, ITKApps failk to build. I narrowed it
>>>>>>>> down using git commit.
>>>>>>>>
>>>>>>>> http://review.source.kitware.com/#/c/19729/
>>>>>>>>
>>>>>>>> Here is the failing dashboard entry:
>>>>>>>> https://open.cdash.org/viewBuildError.php?buildid=3828894
>>>>>>>>
>>>>>>>> I'll take a lot to see if I can fix it.
>>>>>>>>
>>>>>>>> Bill
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Unpaid intern in BillsBasement at noware dot com
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Unpaid intern in BillsBasement at noware dot com
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Unpaid intern in BillsBasement at noware dot com
>>>>> _______________________________________________
>>>>> Powered by www.kitware.com
>>>>>
>>>>> Visit other Kitware open-source projects at
>>>>> http://www.kitware.com/opensource/opensource.html
>>>>>
>>>>> Kitware offers ITK Training Courses, for more information visit:
>>>>> http://kitware.com/products/protraining.php
>>>>>
>>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>>
>>>>> Follow this link to subscribe/unsubscribe:
>>>>> http://public.kitware.com/mailman/listinfo/insight-developers
>>>
>>>
>>>
>>> --
>>> Unpaid intern in BillsBasement at noware dot com
>>> _______________________________________________
>>> Powered by www.kitware.com
>>>
>>> Visit other Kitware open-source projects at
>>> http://www.kitware.com/opensource/opensource.html
>>>
>>> Kitware offers ITK Training Courses, for more information visit:
>>> http://kitware.com/products/protraining.php
>>>
>>> Please keep messages on-topic and check the ITK FAQ at:
>>> http://www.itk.org/Wiki/ITK_FAQ
>>>
>>> Follow this link to subscribe/unsubscribe:
>>> http://public.kitware.com/mailman/listinfo/insight-developers
>>> _______________________________________________
>>> Community mailing list
>>> Community at itk.org
>>> http://public.kitware.com/mailman/listinfo/community
>>
>
>
>
> --
> Unpaid intern in BillsBasement at noware dot com
--
Unpaid intern in BillsBasement at noware dot com
More information about the Insight-developers
mailing list