IRC logs for #buildstream for Friday, 2020-01-17

*** phildawson has quit IRC00:13
*** tristan has joined #buildstream02:22
*** dellie__ has joined #buildstream03:32
*** delli3_ has quit IRC03:34
*** narispo has quit IRC03:57
jjardonHi benschubert04:59
jjardondid you tag this release? https://gitlab.com/BuildStream/bst-plugins-experimental/commit/4ebd361daff161308166f4aa845b0c9d41fe784004:59
jjardonI think we should bump the version of bst-plugins-experimental to 0.90.1 or something similar, otherwise the tag is going to conflict with the existing 0.1405:00
jjardon(from the bst-external, ie bst-1)05:00
jjardonbenschubert: actually, maybe 1.90 and make it match current buildstream version maybe05:01
*** lchlan has quit IRC05:33
*** lchlan has joined #buildstream05:33
gitlab-br-botjjardon merged MR !1729 (jjardon/move_deb_source->master: Remove deb surce plugin, it has beem moved to bst-plugins-experimental) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/172906:12
*** traveltissues has joined #buildstream08:33
*** tpollard has joined #buildstream09:00
benschubertjjardon: the tag should not conflict, they are different repositories and packages no? However 0.91 seems a good idea yup. I don't think we really need to match BuildStream's version, since it's anyways going to be exploded later on09:24
*** tme5 has joined #buildstream09:25
*** tiagogomes has joined #buildstream10:01
coldtomcould we get https://gitlab.com/BuildStream/bst-plugins-experimental/merge_requests/69 merged please?10:02
coldtommaybe worth a bug fix release tbh10:02
benschubertcoldtom: cs-shadow was planning in making another breaking  change on plugins today, and doing a release after that. Would you agree in waiting for this for the release?10:05
benschubertFeel free to merge this change, thanks for the catch (Or do you need me to merge?)10:05
coldtomi don't mind waiting for the release if there's one coming today :P i can't merge myself i'm afraid10:06
benschubertSet to merge! Ok, I'll publish a new one if cs-shadow doesn't has time to finish it today10:07
coldtomta benchubert!10:13
*** jonathanmaw has joined #buildstream10:38
*** cs-shadow has joined #buildstream10:50
*** rdale has quit IRC11:02
coldtomwe should probably add black to bst-plugins-experimental too, shouldn't we11:16
benschubertI'd be for ti :)11:16
benschubert*it11:16
benschubertwould have definitely caught that11:16
* coldtom wants to stop thinking about wrapping his lines11:17
gitlab-br-botcs-shadow opened MR !1803 (chandan/plugin-namespace-fix->master: Plugins: Update entrypoint group for pip origin) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180311:33
coldtomnice change cs-shadow \o/11:34
cs-shadow:)11:36
*** rdale has joined #buildstream11:47
gitlab-br-botcs-shadow opened (was WIP) MR !1803 (chandan/plugin-namespace-fix->master: Plugins: Update entrypoint group for pip origin) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180311:51
*** santi has joined #buildstream11:51
*** santi has quit IRC12:10
*** santi has joined #buildstream12:16
coldtomerm, it seems like it's impossible to write some kinds of plugin tests without using internal buildstream api, is this a known problem?12:37
coldtome.g. to check a plugin throws the correct configurable failure properly, I seem to need to get ErrorDomain from _exceptions12:38
coldtomseveral plugin tests also generate little projects using the yaml dumper from _yaml12:39
*** delli3_ has joined #buildstream13:14
gitlab-br-botcs-shadow opened (was WIP) MR !1803 (chandan/plugin-namespace-fix->master: Plugins: Update entrypoint group for pip origin) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180313:16
*** dellie__ has quit IRC13:17
*** brlogger has joined #buildstream13:23
gitlab-br-bottpollard approved MR !1802 (juerg/assemble_done->master: element.py: Optimize assemble_done()) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180213:32
benschubertcoldtom: would exposing them in 'testutils' do the trick for you? I think it would be fine with making them more open for this purpose (but not for general consumption in the plugin itself)13:54
*** traveltissues has quit IRC13:56
coldtombenschubert: is there a way to import from bst testutils in external plugin repos?14:19
benschubertcoldtom: if it is in buildstream.testutils yes. If it is under tests/tesutils no14:20
coldtomah, yes i'd be happy for it to be in buildstream.testing (especially ErrorDomain as public API there expects that as an arg)14:21
benschubertcoldtom: would you mind making either an issue with the list of things or directly a PR? :) Or maybe a ML post? I don't know what's best there?14:22
coldtombenschubert, sure! i'll need to take a closer look at what's being used first14:26
*** traveltissues has joined #buildstream14:37
coldtomhmm, i think we can probably limit the _yaml surface exposed by just giving a wrapper to _yaml.roundtrip _dump14:49
benschubertcoldtom: for what are you using it generally? Creating elements? We could have a 'write_element' method exposed14:54
benschubertIf you need if for more, we could see to have a nicer way :)14:55
coldtombenschubert:  the existing tests seem to use it to dump a template project.conf14:55
benschubertcreate_project() and create_element() ?14:56
coldtoma fair few even implement methods like that themselves14:56
benschubertMight be a good time to extract and consolidate some of those operations if you have ideas, please open either issues or MRs :D14:57
gitlab-br-botBenjaminSchubert approved MR !1803 (chandan/plugin-namespace-fix->master: Plugins: Update entrypoint group for pip origin) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180315:34
gitlab-br-botcs-shadow merged MR !1803 (chandan/plugin-namespace-fix->master: Plugins: Update entrypoint group for pip origin) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180316:41
*** santi has quit IRC16:53
*** santi has joined #buildstream16:53
*** tpollard has quit IRC17:01
gitlab-br-botcs-shadow opened MR !1804 (chandan/1.93.0->master: NEWS: Add mark for 1.93.0) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180417:02
*** santi has quit IRC17:19
*** tme5 has quit IRC17:23
gitlab-br-botBenjaminSchubert approved MR !1804 (chandan/1.93.0->master: NEWS: Add mark for 1.93.0) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180417:31
*** santi has joined #buildstream17:50
gitlab-br-botcs-shadow merged MR !1804 (chandan/1.93.0->master: NEWS: Add mark for 1.93.0) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180417:52
*** phildawson has joined #buildstream18:06
*** phildawson has quit IRC18:09
*** phildawson has joined #buildstream18:09
*** tiagogomes has quit IRC18:12
*** jonathanmaw has quit IRC18:44
*** phildawson has quit IRC19:05
*** santi has quit IRC20:21
*** santi has joined #buildstream20:21
coldtomanyone got anything against a backport of !1694 to bst-1? I've opened https://gitlab.com/buildstream/bst-external/merge_requests/114 to backport the bazel_* plugins, but i just realised I relied on having /dev/shm in the sandbox20:30
gitlab-br-botMR !1694: _sandboxbwrap.py: Create /dev/shm in the sandbox https://gitlab.com/BuildStream/buildstream/merge_requests/169420:30
coldtomif i can't get a backport, then i can just mark_directory /dev/shm, which isn't as neat but won't be too much effort20:30
*** cs-shadow has quit IRC20:39
*** santi has quit IRC20:50
coldtomerm, i also feel like we should backport the fix to make bst work with python 3.8 to bst-121:03
*** phildawson has joined #buildstream21:05
gitlab-br-botcoldtom opened MR !1805 (coldtom/backport-dev-shm->bst-1: Backport creation of /dev/shm in the sandbox to bst-1) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/180521:29
coldtomif ^ that is not acceptable to get merged, then https://gitlab.com/BuildStream/bst-external/merge_requests/114 can already deal without it21:52
*** phildawson has quit IRC22:19
*** phildawson has joined #buildstream22:34
*** phildawson has quit IRC23:23

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!