| *** toscalix has quit IRC | 00:16 | |
| *** Prince781 has joined #buildstream | 01:03 | |
| *** mcatanzaro has quit IRC | 03:35 | |
| *** tristan has joined #buildstream | 06:22 | |
| *** Prince781 has quit IRC | 08:44 | |
| *** valentind has joined #buildstream | 09:10 | |
| *** dominic has joined #buildstream | 09:22 | |
| *** dominic has left #buildstream | 09:22 | |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: WIP: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 09:28 |
|---|---|---|
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: WIP: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 09:28 |
| jmac | ssssam[m]: Aha! Thanks for the advice. | 09:35 |
| *** adds68 has joined #buildstream | 09:36 | |
| *** adds68 has quit IRC | 09:37 | |
| *** aday has joined #buildstream | 09:38 | |
| *** jonathanmaw has joined #buildstream | 09:44 | |
| *** adds68 has joined #buildstream | 10:02 | |
| jonathanmaw | Hi tristan, I've been told there's a request for buildstream to be able to handle multiple repo aliases, but I can't find any open issue that discusses it | 10:08 |
| jonathanmaw | the closest I could find was https://gitlab.com/BuildStream/buildstream/issues/179 and that's only tangentially related | 10:09 |
| tristan | that doesnt seem related | 10:12 |
| tristan | that is artifact cache related, source aliases are different | 10:12 |
| tristan | there was a request I think to do it in `git.py` plugin specifically, a weird request | 10:13 |
| tristan | something about having a single source declaration have fallback uris to try the next one and then the next one | 10:13 |
| tristan | jonathanmaw, that issue I rejected in favor of proper source mirroring support, because the original request is really not desirable | 10:13 |
| tristan | maybe this is not the same request you are talking about, but it seems to me that multiple repo aliases falls into the category of "Think about doing source mirroring properly" | 10:14 |
| jonathanmaw | okay, so I should put together my own gitlab issue for this. | 10:15 |
| tristan | Or, maybe you should A.) find out the reason *why* this is being requested, and B.) figure out if that would *also* be solved by having support for source mirroring | 10:15 |
| tristan | and then C.) Write up a proposal to the list for how you might think source mirroring should work | 10:16 |
| juergbi | jonathanmaw, tristan: https://gitlab.com/BuildStream/buildstream/issues/244 | 10:17 |
| tristan | Ah right, there it is :) | 10:18 |
| *** valentind has quit IRC | 10:28 | |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: WIP: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 10:34 |
| *** aday has quit IRC | 10:46 | |
| gitlab-br-bot | buildstream: merge request (change_the_location_of_generate-base.sh->master: Changed the location of generate-base.sh) #307 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/307 | 11:09 |
| *** aday has joined #buildstream | 11:24 | |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: WIP: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 11:24 |
| gitlab-br-bot | buildstream: merge request (change_the_location_of_generate-base.sh->master: Changed the location of generate-base.sh) #307 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/307 | 11:26 |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 11:27 |
| tristan | juergbi, so fyi... I've added an `environment` section to the `shell` project configuration, along side the existing `environment-inherit`, in the above MR 306 | 11:29 |
| tristan | I wondered if maybe we should nuke `environment-inherit`, since one can achieve the same with the `environment` dictionary, with more verbosity | 11:29 |
| tristan | but think lets just leave it in as it's a bit more convenient in the typical cases ? | 11:30 |
| tristan | juergbi, any other API concerns before landing this ? | 11:30 |
| juergbi | tristan: if `environment-inherit` didn't exist yet, I probably wouldn't add it but I don't have strong feelings on removing it | 11:34 |
| juergbi | no concerns come to mind right now | 11:35 |
| juergbi | (as it's only needed in project.conf, i.e., once per project, convenience is not the biggest concern for me) | 11:36 |
| tristan | Right, I think we *could* potentially remove before a 1.2 if we want to | 11:36 |
| tristan | If we do remove it, the main advantage I can see is more concise documentation | 11:36 |
| juergbi | yes, the configuration structure would be slightly simpler / stream-lined | 11:36 |
| juergbi | however, it doesn't matter much, imo | 11:37 |
| tristan | Well, maybe I should do it now | 11:37 |
| juergbi | if we remove it, we should probably do it as soon as possible | 11:37 |
| juergbi | to avoid other people having to migrate project.conf | 11:37 |
| tristan | it's not much work, I'll see what the project.conf for gnome-build-meta looks like | 11:37 |
| tristan | and with versioning it's nice that nothing breaks really | 11:37 |
| tristan | yeah, I'm gonna kill it | 11:39 |
| tristan | juergbi, I can see that project.conf is *also* more readable without a mixture of both | 11:39 |
| tristan | that sells me, work required is almost none | 11:40 |
| tristan | I think with this, I'm ready to roll out 1.1.1 | 11:40 |
| tlater | Oooh, another new buildstream release :) | 11:42 |
| jjardon[m] | \o/ | 11:49 |
| jonathanmaw | \o/ | 11:50 |
| jonathanmaw | oh, one thing, I think | 11:51 |
| tristan | yeah this one mostly marks the addition of shell features making testing with `bst shell` more useful | 11:51 |
| jonathanmaw | tristan: should we slip in https://gitlab.com/BuildStream/buildstream/merge_requests/304 first? | 11:51 |
| tristan | jonathanmaw, I'm not picky about that right now to be honest | 11:59 |
| jonathanmaw | okie doke | 11:59 |
| tristan | jonathanmaw, mostly because, people usually read the docs on the web page, I dont think anyone is building it themselves | 11:59 |
| tristan | jonathanmaw, wait a sec... I *AM* certainly picky about that, just dont think it merits holding the release for :D | 12:00 |
| jonathanmaw | heh | 12:00 |
| tlater | tristan: This is entirely unrelated, but should we perhaps start having a look at opening the testutils api? jennis_ has had to copy that module to bst-external, it would be nice to nip that duplication in the bud. | 12:01 |
| tristan | tlater, I've mentioned before that I'm interested in sharing this API yes, the only question is who will work on doing it :) | 12:01 |
| tristan | and when :) | 12:02 |
| tlater | Do we have an open issue yet? | 12:02 |
| * tlater can't remember off the top of his head | 12:02 | |
| tristan | tlater, you will also see that I've done a minor refactor in the integration Cli fixture | 12:02 |
| tristan | after I merge my branch... | 12:02 |
| * tlater opens an issue for opening the testutils api - probably more likely we'll get someone to do it that way ;) | 12:04 | |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 12:05 |
| tlater | tristan: Ah, I see, yeah, that's prettier :) | 12:07 |
| tlater | Better than trying to force the same project.conf on everything | 12:07 |
| tristan | tlater, yeah this is a hybrid of both; we keep a template around where the source aliases are defined in one place; and composite test case specific stuff on top of the loaded project.conf | 12:09 |
| tristan | api is better, and stuff defined in less places | 12:09 |
| gitlab-br-bot | buildstream: issue #281 ("Make the testutils API public") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/281 | 12:20 |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: Support env var expansion in shell configuration) #306 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 12:20 |
| gitlab-br-bot | buildstream: merge request (jmac/build-uid->master: WIP: Specify custom UID for build sandbox in elements) #301 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/301 | 12:24 |
| ltu | anyone have anything to discuss for the meeting at 14:00 UTC? | 12:27 |
| tristan | I was going to ask the same :) | 12:30 |
| tristan | agenda looks thin, and at the same time things are pretty fast moving | 12:31 |
| gitlab-br-bot | buildstream: merge request (239-use-pylint-for-linting->master: Resolve "Use pylint for linting") #283 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/283 | 12:33 |
| *** mcatanzaro has joined #buildstream | 12:35 | |
| gitlab-br-bot | buildstream: merge request (tristan/shell-enhancements->master: Support env var expansion in shell configuration) #306 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/306 | 12:37 |
| *** aday has quit IRC | 12:46 | |
| *** ChanServ sets mode: +o tristan | 12:53 | |
| *** tristan changes topic to "/BuildStream 1.1.1 is out ! | https://gitlab.com/BuildStream/buildstream | Docs: https://buildstream.gitlab.io/buildstream | IRC logs: https://irclogs.baserock.org/buildstream | Roadmap: https://wiki.gnome.org/Projects/BuildStream/Roadmap" | 12:53 | |
| gitlab-br-bot | buildstream: merge request (239-use-pylint-for-linting->master: Resolve "Use pylint for linting") #283 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/283 | 12:55 |
| gitlab-br-bot | buildstream: merge request (239-use-pylint-for-linting->master: Resolve "Use pylint for linting") #283 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/283 | 12:57 |
| jennis_ | juergbi ^^ rebased the linting branch so it's aligned with master | 12:59 |
| juergbi | ta | 13:00 |
| *** jmac has quit IRC | 13:02 | |
| *** jmac has joined #buildstream | 13:02 | |
| gitlab-br-bot | buildstream: issue #223 ("Make it possible to run functional GNOME apps from bst shell") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/223 | 13:07 |
| juergbi | \o/ | 13:08 |
| tristan | another one bites the dust :) | 13:12 |
| * tlater is a bit disappointed at how much code he has to duplicate from OSTree.diff_dirs :| | 13:12 | |
| tlater | Looks like about 200 loc in total | 13:13 |
| *** aday has joined #buildstream | 13:14 | |
| tristan | juergbi, jonathanmaw... just pushed this for docs now, fyi: https://gitlab.com/BuildStream/buildstream/commit/5a0f56af779c18eee27d1ae2d617faa4b01f1687 | 13:14 |
| tristan | that is a pattern now, anything added in the format should indicate the since version | 13:14 |
| juergbi | makes sense, ta | 13:15 |
| * tristan also notes ... | 13:20 | |
| tristan | It's very pleasant how tab completion is much more responsive now :) | 13:20 |
| tristan | noticeably | 13:20 |
| juergbi | yes, i noticed this as well | 13:21 |
| tlater | \o/ | 13:21 |
| tristan | ltu, I'd like to add to the agenda: "Feature prioritizations" | 13:31 |
| *** sstriker has joined #buildstream | 13:33 | |
| tristan | ltu, I have a couple of items, and think it might be good to have some chat by interested parties regarding prioritization | 13:33 |
| * sstriker peeks up | 13:33 | |
| gitlab-br-bot | buildstream: issue #282 ("Status bar regression after enhancing log configurability") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/282 | 13:41 |
| ltu | tristan, sure, added | 13:47 |
| *** tristan has quit IRC | 13:48 | |
| *** tristan has joined #buildstream | 13:55 | |
| tristan | nb all: Monthly (or rather 4-weekly) team meeting starting in a few minutes in #buildstream-meetings | 13:57 |
| *** sstriker has quit IRC | 13:58 | |
| *** jennis_ is now known as jennis | 13:58 | |
| *** sstriker has joined #buildstream | 13:59 | |
| ltu | we're just starting the meeting now, over in #buildstream-meetings | 14:03 |
| *** aday has quit IRC | 14:56 | |
| persia | Meeting log at http://meetbot.gnome.org/buildstream-meetings/2018/buildstream-meetings.2018-03-06-14.04.log.html for anyone who didn't make it. | 15:08 |
| *** aday has joined #buildstream | 15:12 | |
| jennis | The file tests/integration/base/generate-base.sh doesn't seem to be "called" anywhere and when manually run it creates an alpine-base.tar.gz file which looks to contain the sysroot of the alpine docker image | 15:33 |
| jennis | To update the integration tests in bst-external I need to have quilt installed in the alpine docker image, tlater has suggested modifying said script to include quilt and then point the import element to the generated tar file | 15:34 |
| jennis | However, when pointing to this tar file before I make any changes to the .sh script, the majority of integration tests fail | 15:36 |
| jennis | Is this because the use of this tar file is redundant? | 15:36 |
| tristan | The tarball is not to be revisioned in the BuildStream git repo. | 15:36 |
| tristan | Integration tests require an external runtime to be downloaded; it's up to us to host it somewhere if we want to run the integration tests | 15:37 |
| nexus | this whole tar docker img thing is confusing | 15:37 |
| tristan | The script is important if we ever want to be able to update the tarball | 15:37 |
| tristan | Probably the first thing you want to do, is add buildstream/tests/integration/README | 15:38 |
| jennis | I think tlater wads going to add something to HACKING...? | 15:39 |
| tlater | jennis: Yep, not related to this image though; I was only made aware of it by you | 15:39 |
| tristan | HACKING is a good place for this to be documented too yes | 15:40 |
| tlater | Could also document the update process though, I'm pretty sure I know how it works at this point | 15:40 |
| tlater | As for your issue, jennis, I'm pretty sure you are just misconfiguring your base-alpine.bst | 15:40 |
| tlater | Would you mind sharing it so we can have a look? | 15:40 |
| jennis | tristan, the tar file created by this script != https://gnome7.codethink.co.uk/tarballs/integration-tests-base.v1.x86_64.tar.xz | 15:40 |
| tlater | A paste of the test failures would be neat too. | 15:40 |
| jennis | How do you update the latter...? | 15:40 |
| tristan | jennis, first of all; I dont think you update the latter | 15:41 |
| tristan | jennis, if this is something for bst-external, then bst-external uses a different tarball, that is slightly more heavy by virtue of including quilt | 15:41 |
| tristan | next, you figure out hmmm, where should we host this | 15:42 |
| jennis | it does or it should? | 15:42 |
| tristan | bst-external != buildstream | 15:42 |
| tristan | To the maintainers of bst-external: We do not make any promises that our own integration testing sysroot provides things your integration tests depend on. | 15:43 |
| tlater | Right, wouldn't make sense to slow buildstream's tests down for this... | 15:44 |
| jmac | I'm not sure who the maintainers of bst-external are now ssssam's wondered off. | 15:44 |
| tristan | Right, that is a separate topic, I'm just clarifying separation :) | 15:45 |
| tlater | jmac: I think jonathanmaw is practically the maintainer at this point | 15:45 |
| tristan | Also, you might coordinate with https://gitlab.com/BuildStream/buildstream-docker-images | 15:45 |
| tristan | if you want centralization of a common generator script which generates both | 15:45 |
| tristan | and you may want to push for that repo to add automation of hosting of tarballs for both projects in some way | 15:46 |
| tlater | tristan: Would we want to host those tarballs on gnome7? | 15:46 |
| tlater | dockerhub can't host tarballs unfortunately | 15:46 |
| tristan | that might be a reasonable proposal, that someone might want to write up, if they wanted such separation, because of something they are working on. | 15:47 |
| jennis | tlater, there's my base-alpine.bst | 15:47 |
| jennis | tristan, you mentioned that bst-external uses a different tarball | 15:47 |
| tristan | tlater, that does not really mean that https://gitlab.com/BuildStream/buildstream-docker-images doesnt have a .gitlab-ci.yml for automation of "things" | 15:47 |
| jennis | The old style tests do not seem to point to any tarball | 15:48 |
| tristan | jennis, I am not aware of the details of what bst-external uses; I made a statement that buildstream will not guarantee anything to external consumers, when it comes to *our* internal testing sysroot tarball. | 15:48 |
| tlater | jennis: As for what bst-external *currently* uses, it uses a slightly odd version of gnome's flatpak sdk | 15:49 |
| tristan | tlater, I dont know if gnome7 is going to continue being that place we use for that honestly; we may decide to move the sysroot tarball somewhere else, too | 15:49 |
| tristan | tlater, in light of new infra which has arisen for ostree sysroots and artifact caches | 15:50 |
| tlater | Ah, I hadn't heard of that yet | 15:50 |
| tlater | But anyway I think it's important to streamline this process a bit; having someone manually scp a file to gnome7 isn't very newcomer friendly | 15:50 |
| tristan | well, we had some issues with infra stability; and require more production support for artifact caches, than what could immediately be provided for those aarch64 sleds | 15:51 |
| tristan | So gnome7 might go away and become a flathub arm machine, or an aarch64 SDK builder, possibly by way of running a gitlab worker | 15:51 |
| tristan | tlater, exactly, that's the right thinking; probably someone should think it over, come up with a plan and propose it in some way | 15:53 |
| tlater | Assuming we can figure out where to host this I can probably have a look. Wouldn't want poor jennis' branch to be left in the void for too long ;) | 15:54 |
| tristan | One can always come up with a prototype and move to a more permanent hosting later | 15:55 |
| tristan | This is why buildstream has a `sysroot` alias for integration tests, and the gnome7 URL is only specified in exactly one place | 15:56 |
| tristan | (for the alpine sysroot) | 15:56 |
| tlater | Yeah, that seems fine. Man this alias feature is a lot cooler than I originally thought :) | 15:57 |
| gitlab-br-bot | buildstream: merge request (issue-243_dpkg_import_source_plugin->master: WIP: Created DPKG Source plugin for Issue #10) #305 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/305 | 16:29 |
| adds68 | Hi, i have a dependency in a file, that i only want to build on specific platforms, is there anyway to specify this? | 16:40 |
| juergbi | adds68: if you have a project option for this you can use a (?) conditional around the dependency | 16:45 |
| adds68 | Hi juergbi, do these options have to be defined in the project.conf? | 16:48 |
| juergbi | yes | 16:48 |
| juergbi | https://buildstream.gitlab.io/buildstream/projectconf.html#options | 16:48 |
| adds68 | cheers, i shall take a look, i think we do have some options defined | 16:49 |
| *** noisecell has quit IRC | 16:53 | |
| ltu | Milestone created to capture immediate priorities - https://gitlab.com/BuildStream/buildstream/milestones/1 | 17:05 |
| ltu | No dates assigned, for now. | 17:05 |
| tlater | Oh, that means we'll get pretty work-efficiency graphs in the nearish future :) | 17:09 |
| juergbi | only if we added a due date | 17:12 |
| juergbi | and as we use this only to track the immediate priorities, it wouldn't be that useful | 17:12 |
| *** xjuan has joined #buildstream | 17:17 | |
| *** jonathanmaw has quit IRC | 17:19 | |
| *** valentind has joined #buildstream | 18:20 | |
| *** Prince781 has joined #buildstream | 19:47 | |
| *** xjuan has quit IRC | 19:56 | |
| *** Prince781 has quit IRC | 20:16 | |
| *** aday has quit IRC | 20:58 | |
| *** valentind has quit IRC | 21:53 | |
| *** tristan has quit IRC | 22:25 | |
| *** xjuan has joined #buildstream | 22:48 | |
| *** xjuan has quit IRC | 23:00 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!