[Interest] Is moc obsolete?

[Konstantin Tokarev]

08.07.2015, 10:59, "André Somers" <andre at familiesomers.nl>:
> Igor Mironchik schreef op 8-7-2015 om 09:40:
>> Hi,
>>
>> On 08.07.2015 10:12, Dmitry Volosnykh wrote:
>>> Igor, there was quite a hot discussion of this project in development mailing list. Follow here: http://lists.qt-project.org/pipermail/development/2015-June/022164.html
>>
>> Interesting, but I didn't find answer to my main question - about Qt plans to remove moc...
> No, there are no plans. But it is good that people are experimenting with different alternative solutions. For the moment, there is no broad enough support on the horizon yet for any of the technologies needed to actually do this, without leaving essential platforms Qt is supported on behind. That means that for now, it is a no-go. But that doesn't mean the issue won't be revisited in the future and experiments on approaches to replace moc give insight in what's really needed and what is feasible and what is not. I, for one, also liked Woboq's experiments to reimplement moc using clang extensions, making the code generation part of the actual compilation. But it is obvious that that is not going to work on, say, MSVS 2012.

Clang-based moc could be run as separate tool (just like current implementation of moc). It would allow to support more C++ features in moc and keep compatibility with all compilers.


-- 
Regards,
Konstantin