IRC logs for #buildstream for Wednesday, 2019-12-18

« Tuesday, 2019-12-17 Index Thursday, 2019-12-19 »
*** narispo has joined #buildstream00:14
*** narispo has quit IRC02:58
*** narispo has joined #buildstream03:11
*** narispo has quit IRC03:37
*** narispo has joined #buildstream03:38
*** narispo has quit IRC04:09
*** narispo has joined #buildstream04:12
*** narispo has quit IRC04:30
*** narispo has joined #buildstream04:32
*** narispo has quit IRC05:20
*** narispo has joined #buildstream05:24
*** toscalix has joined #buildstream08:54
*** toscalix has quit IRC09:14
*** rdale has joined #buildstream09:29
*** tme5 has joined #buildstream09:36
*** toscalix has joined #buildstream09:37
*** traveltissues has joined #buildstream09:40
*** santi has joined #buildstream09:50
tme5i'm having difficulty again understanding the tests in testing/_sourcetests/mirror.pt09:51
tme5.py - as far as I can see there is no documentation on what mirrors really do, and nothing on what they should do in the presence of aliases09:52
juergbitme5: have you seen https://docs.buildstream.build/master/format_project.html#mirrors ?10:05
tme5yes, but it doesn't seem complete10:06
tme5so mirrors define sets of aliases that can be switched between with configuration?10:07
tme5basically i'm just trying to understand what the correct behaviour should be for some of these tests that set an alias both in a mirror and outside one10:07
juergbithe URLs defined for aliases in a mirror are alternatives to the default URL defined by the non-mirror aliases configuration10:08
*** toscalix has quit IRC10:09
juergbiI haven't been involved in that area, so would have to check the details10:10
tme5alright - i understand as well there is some kind of fallback mechanism that will use different aliases between mirrors if one fails?10:10
tme5ah ok, no worries, i appreciate the help10:10
juergbitme5: this might help: https://gitlab.com/BuildStream/buildstream/issues/32810:12
juergbiwe should expand the documentation10:12
tme5juergbi, this is helpful, i think it also explains unintuitive and undocumented behaviour of translate_url10:14
tme5thanks!10:14
*** lachlan has joined #buildstream10:26
*** lachlan has quit IRC10:30
*** lachlan has joined #buildstream10:31
*** lachlan has quit IRC10:41
*** lachlan has joined #buildstream10:45
*** toscalix has joined #buildstream10:45
*** toscalix has quit IRC10:46
*** lachlan has quit IRC10:51
*** narispo has quit IRC10:56
*** narispo has joined #buildstream11:03
*** toscalix has joined #buildstream11:24
*** tpollard has quit IRC11:24
*** tpollard has joined #buildstream11:24
*** toscalix has quit IRC11:35
*** toscalix has joined #buildstream11:37
*** toscalix has quit IRC11:40
*** lachlan has joined #buildstream11:51
*** lachlan has quit IRC12:00
*** lachlan has joined #buildstream12:07
*** lachlan has quit IRC12:11
*** lachlan has joined #buildstream12:16
*** lachlan has quit IRC12:26
*** lachlan has joined #buildstream12:29
*** santi has quit IRC12:29
*** lachlan has quit IRC12:32
*** gokcennurlu has joined #buildstream12:34
*** gokcennurlu has left #buildstream13:02
*** gokcennurlu has joined #buildstream13:02
*** gokcennurlu has quit IRC13:03
*** gokcennurlu has joined #buildstream13:03
gokcennurluhi all. is there a use case for "bst shell--use-buildtree" (without -b) ?13:12
*** santi has joined #buildstream13:13
gokcennurluI was skimming over https://gitlab.com/BuildStream/buildstream/merge_requests/1767, particularly the help_text for `bst shell --use-buildtree`. I followed the developments in https://gitlab.com/BuildStream/buildstream/merge_requests/986 and the MRs followed, in addition to ML discussions about CLI consistency13:13
tpollardWSalmon might have some thoughts13:17
*** lachlan has joined #buildstream13:21
WSalmonummm, im not sure that i can think of many really good use cases but some times when im developting a app i might want to try the old and new along side eachother particlly for GUI or math style things, but its not the most compelling use case. or the best way to do it. Is it  causing a lot of pain supporting it? surrly the pain comes from being able to do ether, is both much more of a cost? I imaging some one had a good use case at some point and13:27
WSalmonwe have just forgoten13:27
*** lachlan has quit IRC13:28
benschubert^ To be more precise, this is about the build tree in a "run" shell, not the build shell (where it does make sense)13:56
*** slaf has quit IRC14:00
*** slaf has joined #buildstream14:17
*** slaf has joined #buildstream14:17
*** slaf has joined #buildstream14:17
*** slaf has joined #buildstream14:17
tpollardeek, getting a lot of valueerrors whilst targeting the debian target14:30
benschuberttpollard: valuerrrors?14:33
tpollardwill capture the outpout in a sec14:33
tpollardit's related to the process close(), but not because of python version14:33
tme5i've been getting loads of them14:34
*** slaf_ has joined #buildstream14:34
*** slaf_ has joined #buildstream14:34
tme5i thought it was due to race conditions in my plugin changes14:34
tme5tpollard, is it this? "ValueError: Cannot close a process while it is still running. You should first call join() or terminate()."14:35
*** slaf has quit IRC14:35
*** slaf_ is now known as slaf14:35
tpollardit's because we actually try & explicitly close the subprocess now14:35
*** lachlan has joined #buildstream14:36
tpollardhttps://gitlab.com/BuildStream/buildstream/snippets/192356614:37
tpollardhttps://gitlab.com/BuildStream/buildstream/merge_requests/1766/diffs#d7d6127e94f4700edb5c77e9594ff85138d19dcd_453_45314:37
tpollardI don't think we want to be joining here14:42
*** lachlan has quit IRC14:44
tpollardI'll create an issuer14:45
tme5i'm almost certain SourceFetcher.fetch never gets passed the alias_override option. anyone remember if this was removed?14:48
*** lachlan has joined #buildstream14:50
gitlab-br-bottpollard opened issue #1243 (close() on child subprocesses is causing ValueErrors on Python 3.7+) on buildstream https://gitlab.com/BuildStream/buildstream/issues/124314:52
*** lachlan has quit IRC14:56
*** lachlan has joined #buildstream15:03
*** lachlan has quit IRC15:08
*** lachlan has joined #buildstream15:18
WSalmonbenschubert, yep, My point was that some times when developing a a app its nice to have the installed version along side, and a build shell dose not give you that, that said im not sure a `run` shell with buildtree would give you that ether15:24
benschubertit would not15:24
*** lachlan has quit IRC15:28
*** lachlan has joined #buildstream15:43
*** lachlan has quit IRC15:53
*** lachlan has joined #buildstream15:56
*** lachlan has quit IRC16:00
jjardonbenschubert: is anything else remaining for https://gitlab.com/BuildStream/buildstream/merge_requests/1757 ?16:05
*** bochecha_ has joined #buildstream16:08
*** bochecha has quit IRC16:09
*** bochecha_ is now known as bochecha16:09
benschubertjjardon: nobody answered my ML post, I was hoping to get some approval there :) But I guess I can send a reminder then update and merge16:11
*** lachlan has joined #buildstream16:11
traveltissuesbenschubert, i think i did respond16:11
benschuberttraveltissues: oups you are right, I'll catch up on that16:12
traveltissueshttps://mail.gnome.org/archives/buildstream-list/2019-December/msg00018.html16:12
traveltissuesok16:12
jjardonYou have an approval in the MR :) but yeah that looks like a plan16:12
benschubertjjardon: awesome thanks :)16:13
*** lachlan has quit IRC16:36
*** santi has quit IRC16:43
*** bochecha_ has joined #buildstream16:52
*** bochecha has quit IRC16:55
*** bochecha_ is now known as bochecha16:55
*** lachlan has joined #buildstream16:59
*** santi has joined #buildstream17:01
*** lachlan has quit IRC17:05
*** lachlan has joined #buildstream17:18
*** traveltissues has quit IRC17:20
*** lachlan has quit IRC17:22
tme5i won't be working on buildstream for a few weeks; i've opened an MR for my work so far as !1774. Feel free to leave any comments on the idea or the work so far17:24
gitlab-br-botMR !1774: WIP: Merge 'git_tag' functionality into 'git' source https://gitlab.com/BuildStream/buildstream/merge_requests/177417:24
benschuberttme5: might be good to open a discussion on the ML for this17:26
tme5alright, i'll do so when i'm back. happy holidays all17:28
*** lachlan has joined #buildstream17:30
*** tme5 has quit IRC17:31
benschubertyou too!17:33
*** lachlan has quit IRC17:34
*** bochecha has quit IRC17:38
*** phoenix has joined #buildstream17:47
*** lachlan has joined #buildstream17:49
*** lachlan has quit IRC17:54
*** tiagogomes has quit IRC18:09
gitlab-br-botBenjaminSchubert opened issue #1244 (Don't use psutil.Process.cpu_affinity on platform not supporting it) on buildstream https://gitlab.com/BuildStream/buildstream/issues/124418:10
*** santi has quit IRC18:10
*** rdale has quit IRC18:22
*** narispo has quit IRC18:35
*** phoenix has quit IRC18:45
*** benschubert has quit IRC20:40
*** gokcennurlu has quit IRC23:05
« Tuesday, 2019-12-17 Index Thursday, 2019-12-19 »

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