| *** mohan43u has quit IRC | 00:28 | |
| *** mohan43u has joined #buildstream | 00:32 | |
| *** xjuan has quit IRC | 01:02 | |
| *** alatiera has joined #buildstream | 01:12 | |
| *** alatiera has quit IRC | 01:17 | |
| *** kapil___ has joined #buildstream | 02:01 | |
| *** juergbi has quit IRC | 03:16 | |
| *** juergbi has joined #buildstream | 03:25 | |
| *** nimish has joined #buildstream | 04:11 | |
| *** kapil___ has quit IRC | 04:50 | |
| *** nimish has quit IRC | 05:32 | |
| *** toscalix has joined #buildstream | 06:44 | |
| *** nimish has joined #buildstream | 07:06 | |
| *** tristan has joined #buildstream | 07:07 | |
| *** juergbi has quit IRC | 07:08 | |
| *** lagarcia_ has joined #buildstream | 07:09 | |
| *** juergbi has joined #buildstream | 07:10 | |
| *** tristan has quit IRC | 08:07 | |
| *** tristan has joined #buildstream | 08:33 | |
| *** nimish has quit IRC | 08:36 | |
| *** nimish has joined #buildstream | 08:37 | |
| *** nimish has quit IRC | 08:47 | |
| *** nimish has joined #buildstream | 08:47 | |
| *** nimish has quit IRC | 09:02 | |
| *** nimish has joined #buildstream | 09:02 | |
| *** toscalix has quit IRC | 09:08 | |
| *** toscalix has joined #buildstream | 09:10 | |
| *** toscalix has quit IRC | 09:12 | |
| *** kapil___ has joined #buildstream | 09:18 | |
| *** ChanServ sets mode: +o tristan | 09:38 | |
| tristan | juergbi, is there any reason why manual plugin does not set BST_VIRTUAL_DIRECTORY = True ? | 09:38 |
|---|---|---|
| * tristan thinks he will make sure all build elements do this | 09:39 | |
| tristan | looks like there are a bunch that dont, makemaker does not, and make does; very strange | 09:40 |
| Kinnison | tristan: It's possibly worth waiting to see if jmac can weigh in on this, I think BST_VIRTUAL_DIRECTORY was from his introduction of the CASCache stuff ready for remote execution | 09:42 |
| tristan | Kinnison, it was python and perl related plugins that were left behind, makemaker, modulebuild, pip and distutils | 09:45 |
| tristan | Oh, and manual | 09:45 |
| tristan | They are all simple BuildElement implementations, and don't behave any differently except for having different default commands | 09:46 |
| gitlab-br-bot | tristanvb opened MR !1009 (tristan/dont-batch-prepare-assemble-by-default->master: Dont batch prepare assemble by default) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1009 | 09:54 |
| *** jonathanmaw has joined #buildstream | 09:56 | |
| *** tiagogomes has joined #buildstream | 09:59 | |
| *** raoul has joined #buildstream | 10:01 | |
| *** nimish has quit IRC | 10:02 | |
| *** nimish has joined #buildstream | 10:03 | |
| benschubert | Hey tristan , if you have time, it would be nice if you could give a final check to !702 before I go merge since it uses mocking in tests :) | 10:09 |
| gitlab-br-bot | MR !702: Mock storage space checks for tests. https://gitlab.com/BuildStream/buildstream/merge_requests/702 | 10:09 |
| Nexus | juergbi: re: https://gitlab.com/BuildStream/buildstream/merge_requests/925#note_124450122 Do we actually support having a workspace which is also an project directory? This seems very odd to me | 10:15 |
| Kinnison | can we open a workspace on a junction, to edit the junction? | 10:16 |
| Kinnison | If so, then yes | 10:16 |
| Nexus | apparently we can | 10:19 |
| tristan | benschubert, oh still this one ! | 10:22 |
| benschubert | tristan: Yes, sorry about that, it's the last test that prevents the testsuite from working on my personnal laptop :) | 10:22 |
| tristan | benschubert, lets merge it... needs a rebase | 10:23 |
| * tristan hits button | 10:24 | |
| tristan | benschubert, lucky you, you dont have dozens of failing linter checks ? | 10:24 |
| benschubert | Mmh not last time I checked, but I haven't tested it for a while, been busy with fixing wsl failures :) | 10:29 |
| *** lachlan has joined #buildstream | 10:30 | |
| tristan | benschubert, there is an issue of the linter updates and the linter's dependency updates inadvertently changing the behavior of the linter, such that things work in CI, but fail depending on what you happen to have on your laptop | 10:31 |
| tristan | which is... really annoying | 10:31 |
| tristan | I mean, the linter finding new problems is always good, but random testing behavior is not | 10:32 |
| benschubert | tristan: One thing I used to ud to avoid that is have two files: "dev-requirements.in", containing my base requirements with version range, and then a "dev-requirements.txt" which freeze the exact version of all dependencies. It allows to prevent such random behavior, having developers always install the .txt version, and the other one being used only when upgrading all deps | 10:33 |
| benschubert | Which was propsoed actually by cs-shadow in https://mail.gnome.org/archives/buildstream-list/2018-December/msg00054.html , I guess all those problem will go away once we do this :) | 10:34 |
| tristan | benschubert, I tried a variety of version pinnings but couldnt find exactly the relevant things; of course at least separating the deps for developers and those for runtime is good, but then the point of having multiple CI runs on multiple distros is sort of to ensure that we do still pass our tests on a variety of likely existing dependency versions | 10:37 |
| *** nimish has quit IRC | 10:38 | |
| benschubert | tristan: sorry, I meant the .txt being used in the CI and by developers and the .in verison only used when someone wants to explicitely update the versions :) | 10:39 |
| tristan | jmac, I put 5 commits on this branch which extends support for virtual directories to the build elements where I think it was simply missed: https://gitlab.com/BuildStream/buildstream/merge_requests/1009/commits | 10:39 |
| benschubert | Ah no sorry misread you | 10:39 |
| tristan | benschubert, it's a bit of a pickle yeah | 10:39 |
| tristan | I think for now the first thing is to just fix all the linting problems with the latest versions of things | 10:40 |
| tristan | and anyway the CI remains stable because it always has exact versions of things in the docker images | 10:40 |
| benschubert | We can also use the .in/.txt separation for dev-requirements anyways, since this if for developers and not for users | 10:40 |
| benschubert | This would reduce the problems already | 10:41 |
| tristan | only problem is *sometimes* we need to fix new linting errors on peoples laptops when they upgrade to new versions we havent seen yet | 10:41 |
| *** nimish_ has joined #buildstream | 10:41 | |
| tristan | benschubert, it's kind of desirable to have developers work with a variety of environments too though right ? I mean we should catch more problems that way | 10:41 |
| Nexus | tristan: did you have a chance to look at https://gitlab.com/BuildStream/buildstream/merge_requests/925#note_124513015 ? | 10:43 |
| *** nimish_ has quit IRC | 10:43 | |
| benschubert | tristan: Yes and no, I would expect every developer on a python project to use a virtual environment specific for this package, and having the test dependencies then would not be too disturbing I would say. I do agree on the normal requirements though, since those are the ones that might change for users | 10:43 |
| *** nimish_ has joined #buildstream | 10:43 | |
| tristan | benschubert, yeah I agree as far as dev-only requirements go, since we *can* control those | 10:45 |
| tristan | benschubert, but then there is always inadvertent overlap :) | 10:45 |
| tristan | Nexus, lemme check | 10:45 |
| benschubert | obviously, but we cannot handle everything sadly :) | 10:46 |
| benschubert | On the other hand, having tox would allow us to test multiple versions of the normal dependencies too actually, we could have something like "base minimum versions", "version proven to work (best versions to use)", and "latest versions, allowed to fail", what would you say about that? | 10:46 |
| *** nimish_ has quit IRC | 10:48 | |
| gitlab-br-bot | tristanvb closed issue #530 (Cache quota size definition tests depend on the host system) on buildstream https://gitlab.com/BuildStream/buildstream/issues/530 | 10:58 |
| gitlab-br-bot | tristanvb merged MR !702 (BenjaminSchubert/fix-quota-tests->master: Mock storage space checks for tests.) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/702 | 10:58 |
| benschubert | Thanks a lot :D | 10:59 |
| juergbi | tristan: no, I can't think of a reason why these other plugins wouldn't be compatible with virtual directories. for the manual one there is already a patch in a (not directly related) pending MR | 11:06 |
| juergbi | as I mentioned I prefer something like SimpleBuildElement to avoid the duplication, but your MR looks good to me | 11:07 |
| *** alatiera has joined #buildstream | 11:09 | |
| tristan | juergbi, I think we can do that orthogonally, I tend to agree but don't feel entirely certain | 11:10 |
| juergbi | sure, the MR fixes the issue, we can add it later | 11:10 |
| tristan | juergbi, I feel like the greater concern might be to have a way to disable the batching once you have it enabled | 11:11 |
| tristan | Right now it's a now way out situation, which might happen again | 11:11 |
| tristan | but I suppose there is always a way to add more API for disabling it if needed | 11:11 |
| juergbi | I don't currently see a need for disabling | 11:12 |
| juergbi | unless we were to enable it by default again, but that could break some things | 11:12 |
| tristan | juergbi, well, he had one in #800, where you derive from an abstract class which enables it, reuse a lot of it's functionality but only need to disable that | 11:12 |
| gitlab-br-bot | Issue #800: There is no clear way to disable batching of prepare+assemble in BuildElement https://gitlab.com/BuildStream/buildstream/issues/800 | 11:12 |
| tristan | I mean, another class could start to exist, in somebody's project not even in BuildStream, which other people derive from | 11:13 |
| tristan | Right now we've solved it for BuildElement, though, and in a way that doesnt break anything that might exist | 11:13 |
| tristan | again, it's not a problem we really need to think about until it is raised I guess | 11:14 |
| juergbi | ok, sure | 11:14 |
| *** nimish_ has joined #buildstream | 11:18 | |
| *** nimish_ has quit IRC | 11:28 | |
| *** nimish_ has joined #buildstream | 11:29 | |
| *** lachlan has quit IRC | 11:31 | |
| *** nimish_ has quit IRC | 11:33 | |
| *** nimish_ has joined #buildstream | 11:40 | |
| *** lachlan has joined #buildstream | 11:47 | |
| *** nimish_ has quit IRC | 11:49 | |
| *** nimish_ has joined #buildstream | 11:50 | |
| *** alatiera has quit IRC | 11:51 | |
| *** abderrahim has quit IRC | 11:52 | |
| *** abderrahim has joined #buildstream | 11:53 | |
| *** alatiera has joined #buildstream | 11:54 | |
| *** nimish_ has quit IRC | 11:55 | |
| *** nimish_ has joined #buildstream | 11:59 | |
| *** nimish_ has quit IRC | 12:04 | |
| *** nimish_ has joined #buildstream | 12:08 | |
| *** nimish_ has quit IRC | 12:10 | |
| *** nimish has joined #buildstream | 12:11 | |
| *** nimish has quit IRC | 12:21 | |
| *** lachlan has quit IRC | 12:26 | |
| *** nimish has joined #buildstream | 12:26 | |
| *** lachlan has joined #buildstream | 12:28 | |
| *** nimish has quit IRC | 12:33 | |
| gitlab-br-bot | BenjaminSchubert merged MR !1006 (bschubert/fix-unbound-variable-exception->master: element.py: fix unbounded variable in nested python exception) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1006 | 12:34 |
| *** V0lTure has joined #buildstream | 13:10 | |
| *** jferrero has joined #buildstream | 13:54 | |
| *** phildawson has quit IRC | 14:00 | |
| *** phildawson has joined #buildstream | 14:00 | |
| *** raoul_ has joined #buildstream | 14:00 | |
| *** raoul has quit IRC | 14:01 | |
| *** raoul__ has joined #buildstream | 14:02 | |
| *** raoul_ has quit IRC | 14:03 | |
| gitlab-br-bot | knownexus closed MR !926 (issue_640-Build-All->master: Added default element build all functionality to `bst build`) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/926 | 14:18 |
| *** raoul__ has quit IRC | 14:24 | |
| *** raoul__ has joined #buildstream | 14:31 | |
| *** nimish has joined #buildstream | 14:38 | |
| *** raoul__ has quit IRC | 14:43 | |
| gitlab-br-bot | jmacarthur approved MR !985 (gokcennurlu/remote_url_override_push_error->master: Set ArtifactCache for push/pull correctly when `--remote` is used) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/985 | 14:52 |
| *** nimish has quit IRC | 14:53 | |
| *** nimish has joined #buildstream | 14:54 | |
| gitlab-br-bot | tristanvb closed issue #800 (There is no clear way to disable batching of prepare+assemble in BuildElement) on buildstream https://gitlab.com/BuildStream/buildstream/issues/800 | 14:58 |
| gitlab-br-bot | tristanvb merged MR !1009 (tristan/dont-batch-prepare-assemble-by-default->master: Dont batch prepare assemble by default) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1009 | 14:58 |
| *** nimish has quit IRC | 15:09 | |
| *** nimish has joined #buildstream | 15:09 | |
| *** raoul__ has joined #buildstream | 15:10 | |
| *** nimish has quit IRC | 15:21 | |
| gitlab-br-bot | willsalmon approved MR !973 (aevri/contributing_e2e_tests->master: contributing: more clarity on testing) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/973 | 15:30 |
| benschubert | Could someone add me as a developer on the BuildStream organization? It will be simpler than asking for every repo :) | 15:32 |
| juergbi | sure | 15:37 |
| juergbi | benschubert: done | 15:38 |
| benschubert | Thanks! | 15:41 |
| *** lachlan has quit IRC | 16:04 | |
| *** CEnnis91 has joined #buildstream | 16:17 | |
| *** lachlan has joined #buildstream | 16:20 | |
| *** toscalix has joined #buildstream | 17:02 | |
| gitlab-br-bot | jmacarthur opened (was WIP) MR !952 (raoul/627-RE-instance-config->master: Remote-execution instance configuration support) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/952 | 17:19 |
| *** lachlan has quit IRC | 17:21 | |
| *** lsfranco has joined #buildstream | 17:24 | |
| *** lachlan has joined #buildstream | 17:28 | |
| *** lsfranco has quit IRC | 17:29 | |
| *** lsfranco has joined #buildstream | 17:30 | |
| *** lachlan has quit IRC | 17:35 | |
| *** lachlan has joined #buildstream | 17:39 | |
| *** jonathanmaw has quit IRC | 17:44 | |
| *** tpollard has quit IRC | 17:56 | |
| *** raoul__ has quit IRC | 17:57 | |
| gitlab-br-bot | aevri merged MR !973 (aevri/contributing_e2e_tests->master: contributing: more clarity on testing) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/973 | 18:01 |
| *** lachlan has quit IRC | 18:11 | |
| *** xjuan has joined #buildstream | 18:42 | |
| *** lsfranco has quit IRC | 18:45 | |
| *** lsfranco has joined #buildstream | 18:45 | |
| *** nimish has joined #buildstream | 18:50 | |
| *** nimish has quit IRC | 18:59 | |
| *** kapil___ has quit IRC | 19:31 | |
| *** xjuan has quit IRC | 20:00 | |
| *** toscalix has quit IRC | 20:01 | |
| *** xjuan has joined #buildstream | 20:44 | |
| *** jhed9 has joined #buildstream | 22:22 | |
| *** alatiera has quit IRC | 23:25 | |
| *** tristan has quit IRC | 23:31 | |
| *** manfromafar has joined #buildstream | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!