In addition to the MSVC-based build, please support MinGW-w64 [1] on Windows. This basically means allowing to build LibreOffice with GCC and make, or something similar. This will attract more contributors, even tough official builds continue to be based on MSVC. If you are interested in a full build environment, you may also consider MSYS2 [2]. This is basically MinGW-w64 with Cygwin and pacman. No manual dependencies management, wide number of reusable libraries and modern Unix shell. The point is enlarging the developer community, so just MinGW-w64 would help already. Note that MinGW-w64 and MSYS2 have nothing to do with MinGW.org, so if you had a bad experience with MinGW, it may be very different for these other. [1] http://mingw-w64.sourceforge.net [2] http://sourceforge.net/projects/msys2
No. The build system is complex enough as it is. The last thing we need is to introduce more variability into it.
This is a narrow-minded view. The amount of contributors that would feel attracted would probably overcome the complexity of supporting one additional compiler. Otherwise, you can stop using MSVC and more proudly advertise freedom. This is also a harsh attitude with the community. Many people see LibreOffice as mostly a joke compared to Microsoft Office. I would consider not dismissing such potential help. Not even external volunteers, interested in both implementation and maintenance, were considered.
Believe David's perspective comes from his recent endeavors to bring a functional 64-bit build for Windows online with VS 2013. That is coming to fruition--and that has required MSVC for reliable results and build speed of Windows executables. With build tools and code optimization for native MS Visual Studio C++ builds, there is no reason to torture the projects productive maintainers--with Microsoft now giving away VS 2013 very little traction to be had for further effort on MinGW/GCC builds--and they did try! Of course this is a meritocracy--the code is there in git--you are welcome to have at it.
(In reply to V Stuart Foote from comment #3) > Believe David's perspective comes from his recent endeavors to bring a > functional 64-bit build for Windows online with VS 2013. That is coming to > fruition--and that has required MSVC for reliable results and build speed of > Windows executables. That was David Ostrovsky :-) But it is a good example.
(In reply to Renato Silva from comment #2) > This is a narrow-minded view. Maybe, but it is a realistic view. > The amount of contributors that would feel > attracted would probably overcome the complexity of supporting one > additional compiler. Says you. Personally, I do not remember anyone showing interest in native MinGW build on Windows since the beginning of the project. There was an attempt in openoffice.org times, but it did not came to anything. And it was being done by a single person.