Hey everyone, This a rather old topic which was discussed by majority of active developers in IRC for quite some time now, and was also discussed at the at least couple of Blender Conferences.
There are several major issues with SCons in Blender which were never addressed in the past couple of years at least: - There's no active strong maintainer of the building system, meaning all the required tweaks are done by 1.5 developers who aren't really SCons experts - None of the active developers are using SCons which makes it quite difficult to spot bugs in time (yes, there are bugs caused by building system -- missing build flags changes etc) - Divergence between SCons and CMake became quite huge now, this affects on supported feature sets, stability, performance and so on - There are list of other issues, but they aren't so meaningful to mention here Over all the time we were discussing future of SCons nobody jumped in volunteering to become a _strong_ maintainer (meaning, solving actual problems, not just keeping SCons somewhat usable). There were some other stoppers to really bring this discussion to life -- mainly release building process and buildbot environment. Those stoppers are now addressed and all infrastructure switched to CMake (which is good on it's own, because it's now using much better tested build system). So yes, let's start discussion for real: are there real reasons to still keep semi-supported building system in Blender? With the current state of maintenance there my personal answer is -- no. But we've got quite some active developers to get feedback from still. That being said, i also went ahead and prepared the patch to wipe SCons out. So the ones who wants to review the removal code be welcome to [1] (please keep non-codereview discussion here!) [1] https://developer.blender.org/D1680 -- With best regards, Sergey Sharybin _______________________________________________ Bf-committers mailing list [email protected] http://lists.blender.org/mailman/listinfo/bf-committers
