2.3.x does not compile on Windows

[Pavel Sanda]

On Fri, Apr 19, 2024 at 12:52:06PM +0200, Jean-Marc Lasgouttes wrote:
> Le 19/04/2024 ?? 11:12, Pavel Sanda a écrit :
> >On Fri, Apr 19, 2024 at 11:03:55AM +0200, Kornel Benko wrote:
> >>Looks like not so easy. The changes in are
> >>	development/cmake/modules/LyXMacros.cmake: easy managable
> >>	development/cmake/modules/FindCXX11Compiler.cmake: looks not difficult
> >>	CMakeLists.txt:	Huge
> >
> >I thought the changes wouldn't be done in cmake, but in .cpp code,
> >like getting rid of unary_function reported in the begininng of
> >this thread, which we presumably don't use in 2.4 anymore(?).
> 
> I thought about backporting those, but they do not apply cleanly to 2.3, so
> that there is a risk of breaking code (just because a compiler decides to be
> a PITA).
> 
> Normally, we have correct C++11 code, we should be able to compile it in
> C++11 mode (even C++14 would do).

Well, we are not. What about the terrific option of just taking 2.3.7
win binaries, wrap it with 2.4->2.3 lyx2lyx python scripts, release
that as 2.3.7.2 (for win only) and the rest would go as 2.3.8.

The pool of prospective 2.3.8 win users once 2.4 is out would be negligible anyway.

Pavel