| *** gary_perkins has quit IRC | 01:29 | |
| *** gary_perkins has joined #baserock | 01:30 | |
| *** gtristan has quit IRC | 05:34 | |
| *** gtristan has joined #baserock | 05:59 | |
| *** locallycompact has joined #baserock | 08:34 | |
| *** locallycompact has quit IRC | 09:52 | |
| *** ssam2 has joined #baserock | 10:05 | |
| *** ChanServ sets mode: +v ssam2 | 10:05 | |
| ssam2 | hooray! another step forwards | 10:06 |
|---|---|---|
| ssam2 | https://gitlab.com/baserock/definitions/merge_requests/58/diffs | 10:06 |
| ssam2 | removing the .la files from the base sysroot was needed to make cross-compilation work, for some stupid libtool reason | 10:06 |
| ssam2 | but now we can bootstrap baserock from baserock using buildstream | 10:06 |
| gitlab-br-bot | definitions: merge request (sam/bootstrap-from-baserock-binaries->master: BuildStream builds should bootstrap from Baserock binaries, not the Freedesktop SDK) #58 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/58 | 10:07 |
| ironfoot | ace! | 10:17 |
| ssam2 | reviews of the above would be much appreciated | 10:40 |
| ironfoot | will try to review them today | 10:42 |
| gtristan | ssam2, not that it's a big deal; but there might be something more elegant to do than integration commands here: https://gitlab.com/baserock/definitions/merge_requests/58/diffs#6933fc2ad9e2410f5c931679e292765bbd5a564a_0_17 | 10:44 |
| gtristan | maybe I'm wrong, but I think you could define a split-rules domain explicitly for that and exclude it off the bat | 10:45 |
| gtristan | ssam2, I dont think thats a problem with the patch fwiw, only mentioning it for reference :) | 10:46 |
| ssam2 | yeah, i wondered about a separate 'libtool-garbage` split domain. but i'd rather kill the .la files with fire :-) | 10:48 |
| ssam2 | currently they're explicitly in the 'devel' domain, which might make sense in some cases | 10:48 |
| gtristan | Yeah | 10:50 |
| gtristan | ssam2, note that split rules *can* be defined on a per-element basis | 10:50 |
| gtristan | And also, split rules are totally allowed to overlap | 10:50 |
| gtristan | what happens with them, depends on how they are employed (usually in a compose element) | 10:51 |
| ssam2 | ok, cool | 10:51 |
| ssam2 | yeah, putting gcc's .la files in a devel-libtool domain might make sense | 10:51 |
| gtristan | I'm not trying to be picky, just highlighting what's possible :) | 10:55 |
| *** adds68 has quit IRC | 11:32 | |
| *** adds68 has joined #baserock | 11:33 | |
| *** gtristan has quit IRC | 11:42 | |
| *** locallycompact has joined #baserock | 11:56 | |
| *** locallycompact has quit IRC | 11:57 | |
| *** locallycompact has joined #baserock | 11:57 | |
| *** gtristan has joined #baserock | 12:04 | |
| *** locallycompact has quit IRC | 13:22 | |
| *** locallycompact has joined #baserock | 14:11 | |
| *** tiagogomes has quit IRC | 14:33 | |
| *** locallycompact has quit IRC | 15:50 | |
| *** locallycompact has joined #baserock | 16:03 | |
| * ssam2 bravely trying to remove the absolute horror that is the STAGE2_SYSROOT hacks in the gnu-toolchain bootstrap | 16:37 | |
| ssam2 | we don't need it with buildstream, because the bootstrap is in a fixed sandbox rather than in a tmpdir that moves around for each new build | 16:38 |
| ssam2 | however, the old way does work ... i may yet give up and leave it be :-) | 16:39 |
| ssam2 | `sed -e "s@[gMS]\?crt[1in].o%s@%:getenv(STAGE2_SYSROOT %{libdir}/&)@g"` there's a line of build instructions that nobody is gonna miss ... | 16:40 |
| *** jude_ has quit IRC | 16:47 | |
| *** ssam2 has quit IRC | 18:40 | |
| *** adds68 has quit IRC | 21:35 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!