09:53 <beerman> opt/talloc gets a race condition over here with JOBS=40

09:54 <beerman> adding ((JOBS -= 1)) and adjusting the make call to make -j ${JOBS:-1} solves the issue just fine it seems

11:05 <jue> beerman: why not just make -j1?

11:08 <beerman> because I have those cores, but waf is stupid, so granted, -j1 is a safe bet

11:10 <jue> ok

11:17 <jue> for such a small prog like talloc I wouldn't care about parallel builds ;)

11:20 <beerman> granted, its not making a big dent any way ;)

16:01 <jaeger> I've tweaked git gitd rc script on crux.nu a bit to fix the warning/error stenur reported... anyone mind testing that you can clone a repo from crux.nu via git:// rather than ssh?

16:01 <jaeger> s/git gitd/the gitd/

16:08 <ukky> Just tried cloning crux.nu/ports/core.git via git:// protocol. No issues.

16:11 <jaeger> No warning about /root/.config/git/attributes or whatever it was?

16:14 <ukky> No warnings, just usual messages.

16:18 <ukky> https://bpa.st/6NJA

16:20 <jaeger> sweet, thanks for testing :)

16:20 <ukky> any time

18:31 <jue> beerman: new glslang breaks build of mesa for me, a symlink glslang-> glslangValidator seems to fix the issue

18:31 <jue> https://dpaste.com/4HC6QH8F3

19:12 <beerman> jue, oh, i checked a few but not mesa, I'll look into it

19:41 <farkuhar> Eight years after they first proposed renaming glslangValidator, it finally comes to pass: https://github.com/KhronosGroup/glslang/issues/47 ... Between 2015 and 2019, looks like xorgy became less concerned about the burden on downstream packagers.