|
01:30
kjp left
01:34
kjp joined
02:51
harrow left
03:05
harrow joined
11:55
[Coke] left
13:37
[Coke] joined
14:24
librasteve_ joined
18:12
MasterDuke joined
|
|||
| MasterDuke | timo, patrickb, et al.: do you have any suggestions for github.com/MoarVM/MoarVM/pull/1967 ? | 18:13 | |
| tellable6 | 2026-05-10T23:36:30Z #raku-dev <[Coke]> masterduke snyk invite sent | ||
| MasterDuke | [Coke]: thanks | ||
| that PR will let us remove dyncall, which will get our clang CI jobs green again | 18:14 | ||
| lizmat | testing on that PR is pretty red though? | 18:25 | |
| MasterDuke | well, i have every build system i've ever used. it works locally (i.e., MoarVM+NQP+Rakudu all build), but there's a chicken-and-egg problem | 18:29 | |
| s/have/hate/ | |||
| i think the details are if you start from scratch, you don't have libffi.h. moar depends on libffi in the Makefile, but i believe that's only for the final stages? so it starts to compile moarvm, but can't find libffi.h | 18:32 | ||
| and it's just a race condition if you use `-j` | |||
| lizmat | ok, fair enough... but we're close to the release, so [Coke] should really be in on this | 18:34 | |
| MasterDuke | configuring with `--has-libffi` might work, since that might assume there's a system libffi.h available | ||
| no particular urgency, i wouldn't try for this next release | 18:35 | ||
| just pinging people when i happen to be on | |||
| [Coke] | if the build doesn't work with -j it's broken | 18:36 | |
| MasterDuke | i'll be on and off, but will check logs if there are any suggestions | 18:43 | |
|
19:08
MasterDuke left
20:00
rakkable left
20:01
rakkable joined
22:08
ShimmerFairy left,
ShimmerFairy joined
|
|||