Who wants faster LLVM/Clang builds?

I'm sorry I just diving in the maillist but wanna to make a small noise(or joking) here.

Did you tested for all #define branches/cases to prove the tweak is correct?

And I guess no C++ programmer actually cares about the build time here. It's a historic issue. If they do, why they don't go straight for Java?

I'm sorry I just diving in the maillist but wanna to make a small noise(or joking) here.

Did you tested for all #define branches/cases to prove the tweak is correct?

No. Moreover, I know that the tweak is incorrect in some cases, as it currently would wipe all includes from inactive ifdef sections (see “Methodology” in the end of the original email).

And I guess no C++ programmer actually cares about the build time here. It's a historic issue.

If you multiply the number of active developers by an average number of builds they do daily, you may be surprised to see how beneficial it can be to improve build times.

If they do, why they don't go straight for Java?

:slight_smile:

Michael