4 ms·
Actually not true. Many large projects continue to use SCons and new projects as well. I don't think in 2018 we had our heads in the sand about SCons' performan
by onebaddog 6y ago
Actually not true. Many large projects continue to use SCons and new projects as well.
I don't think in 2018 we had our heads in the sand about SCons' performance.
As project maintainer I've (along with other community members) have been working on performance since at least 2016.
- ncmncm 6y agoIf you had a hand in getting it to generate ninja scripts: Congratulations. That made a big improvement in the daily experience of people maintaining projects built under SCons. To be clear: SCons is not notably worse at describing what must be done to build a big project than other systems of its time. Lots of code has been compiled under it. Its deep failure has been in trying to direct the activities of compilations in real time, at which it is bad, gmake is good, and ninja positively excels. Still, slowness in resolving dependencies is hard to defend. It is a problem in operations long well-understood, with familiar algorithms. A data structure well tuned to the algorithm would make even Python able to solve it quickly enough. Nowadays the difficult part of a build system is the complexities of managing foreign dependencies, at which CMake is only just adequate.