| *** mohan43u has quit IRC | 05:47 | |
| *** tristan has quit IRC | 05:53 | |
| *** mohan43u has joined #buildstream | 06:12 | |
| *** tristan has joined #buildstream | 06:16 | |
| *** tristan has quit IRC | 07:38 | |
| *** tristan has joined #buildstream | 07:38 | |
| *** toscalix has joined #buildstream | 07:39 | |
| *** toscalix has quit IRC | 07:40 | |
| *** bochecha has joined #buildstream | 07:49 | |
| *** toscalix has joined #buildstream | 08:05 | |
| *** raoul has joined #buildstream | 08:39 | |
| *** finn has joined #buildstream | 08:47 | |
| gitlab-br-bot | jennis approved MR !1362 (raoul/1024-artifact-docs->master: Update docs regarding artifact and source caches) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1362 | 09:09 |
|---|---|---|
| finn | Any reason tar isn't a core plugin for buildstream? | 09:10 |
| jennis | tar source is, tar element isn't | 09:12 |
| finn | how come? | 09:12 |
| jennis | I guess because legacy reasons, we've supported `bst artifact checkout foo.bst --tar foo.tar.gz` for a very long time now | 09:13 |
| jennis | But to actually have the tarball in your build, to be consumed by another element, you need the plugin, and IIRC the freedesktop guys needed this | 09:14 |
| finn | Heh ok... | 09:16 |
| *** jonathanmaw has joined #buildstream | 09:17 | |
| jennis | finn: my guess would be that because we've decided we're going to start migrating plugins from core, it wouldn't make sense to add another plugin to core, but I'm not 100% sure on that | 09:22 |
| finn | Right, yes someone just told me of that. I wasn't aware | 09:23 |
| finn | Thanks! | 09:23 |
| gitlab-br-bot | marge-bot123 closed issue #1024 (Update arch/design docs to reflect move to Proto based artifact services) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1024 | 09:39 |
| gitlab-br-bot | marge-bot123 merged MR !1362 (raoul/1024-artifact-docs->master: Update docs regarding artifact and source caches) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1362 | 09:39 |
| *** lachlan has joined #buildstream | 09:41 | |
| *** rdale has quit IRC | 10:04 | |
| tpollard | have we made any changes recently that would have significantly reduced the size of the integration cache from the test suite? | 10:25 |
| gitlab-br-bot | raoul.hidalgocharman opened (was WIP) MR !1366 (raoul/1025-legacy-remotes->master: Improved handling of legacy remotes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1366 | 10:25 |
| *** ChanServ sets mode: +o tristan | 10:30 | |
| tristan | tpollard, possibly... is there a commit which uses a slimmed down freedesktop-sdk base instead of the alpine image ? I'm not sure if it might have done the opposite though | 10:30 |
| tristan | i.e. as I recall tlater[m] may have added that, but it may be an *additional* runtime and we did not drop alpine | 10:30 |
| tristan | tpollard, depending on what you are looking at as evidence, also remember the accumulative nature of the gitlab caches | 10:31 |
| tristan | so zapping the gitlab cache may result in removing unused cruft from side branches | 10:31 |
| tpollard | this is locally | 10:32 |
| tpollard | testing against an older commit now for my own sanity :) | 10:32 |
| tlater[m] | tristan it hasn't been added yet | 10:33 |
| tlater[m] | And I'd expect it to increase the size if anything... | 10:34 |
| tristan | right | 10:46 |
| gitlab-br-bot | BenjaminSchubert opened MR !1367 (bschubert/cython-disable-linetrace->master: Setup.py: Disable linetrace by default and only enable when using coverage) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1367 | 10:59 |
| *** alatiera has joined #buildstream | 11:12 | |
| *** lachlan has quit IRC | 11:36 | |
| *** lachlan has joined #buildstream | 11:45 | |
| *** alatiera has quit IRC | 12:27 | |
| *** raoul_ has joined #buildstream | 12:39 | |
| *** raoul has quit IRC | 12:41 | |
| *** bochecha has quit IRC | 12:53 | |
| *** bochecha has joined #buildstream | 12:54 | |
| benschubert | phil: tests/examples/flatpak-autotools: shouldn't be that moved to the ostree plugin tests? | 13:00 |
| tpollard | related https://gitlab.com/BuildStream/buildstream/issues/1032 | 13:02 |
| benschubert | oh thanks tpollard | 13:03 |
| benschubert | Is there any reason why we would want this test here and not in ostree? | 13:03 |
| WSalmon | The examples did/do cover a few diffrent ways to "bootstrap" your project eg, ostree/tar from alphine/fd-sdk, so it is "right" that the tests that cover what the docs do live in examples, but it dosent make sence if they never get run, i think tlater[m] was looking at moving all the examples to use FD-sdk which might solve this too | 13:26 |
| gitlab-br-bot | danielsilverstone-ct approved MR !1367 (bschubert/cython-disable-linetrace->master: Setup.py: Disable linetrace by default and only enable when using coverage) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1367 | 13:27 |
| *** lachlan has quit IRC | 13:28 | |
| tpollard | WSalmon: good points about example test coverage, I hadn't considered that | 13:29 |
| WSalmon | I personally think it would be a shame if the exmples didnt show you at least a variety of ways to boot strap a project after all this is one of the big advantages over yockto/pokey but i think i am a minority voice in this regard, the above from Ben dose seem to be a good reson to move to FD-sdk for all examples to reduce maintance | 13:29 |
| tlater[m] | WSalmon: We're not using ostree to bootstrap anything anymore as far as I am aware, so that would technically obsolete the test case here. | 13:31 |
| tlater[m] | We definitely don't after my branch. Whether we want to keep an example of it for the sake of showing off third party plugins is a different question, of course. | 13:32 |
| tlater[m] | But I think it would make more sense to link to the third party docs in that case. | 13:32 |
| WSalmon | maybe we could just have a exmples stage in the tests taht had lots of plugins installed and then remove the examples from the other tests, this would reduce testing laps-time and total compute time, but still give use a interesting test coverage | 13:33 |
| benschubert | tlater[m]: should I make a PR to the ostree plugin with those tests then? | 13:34 |
| *** lachlan has joined #buildstream | 13:35 | |
| tlater[m] | benschubert: I think that's a good way to go about it, yeah. | 13:36 |
| benschubert | thanks! | 13:36 |
| tlater[m] | As long as we have tests in buildstream that make sure the plugin API stays stable I don't see how this can break it. | 13:37 |
| tlater[m] | benschubert: But please check the docs and make sure the relevant example is also moved (: | 13:37 |
| benschubert | Sure! | 13:38 |
| phil | benschubert, sorry I missed your ping. I don't have much to add to what has already been said, but for what it's worth I agree it makes sense to move that test over. | 13:40 |
| benschubert | thanks phil and no worries! | 13:41 |
| finn | ValueError: Namespace OSTree not available >> What does this mean? | 13:59 |
| finn | Or more importantly, how do I fix it | 14:00 |
| finn | I've installed buildstream and buildstream-externals via pip | 14:00 |
| tpollard | have you added the plugin to your project.conf? | 14:02 |
| jennis | finn: https://docs.buildstream.build/format_project.html#external-plugins | 14:03 |
| jennis | Oh actually, you're using bst 1.2, that's in core | 14:03 |
| jennis | You don't have the right system dependencies, that's the problem | 14:04 |
| finn | jennis, yes as that's what pip provides me | 14:04 |
| finn | I've done an `apt install ostree` | 14:04 |
| finn | What else do I need? | 14:04 |
| jennis | Was distro are you using? | 14:04 |
| jennis | what* | 14:04 |
| finn | ubuntu 18.04 docker image | 14:05 |
| jennis | cool, apt install ostree and pip install pygobject *should* do it | 14:05 |
| jennis | In fact, just the former mayber | 14:05 |
| finn | I already have ostree installed | 14:05 |
| finn | Still doesn't work | 14:07 |
| *** lachlan has quit IRC | 14:07 | |
| juergbi | finn: do you also have gir1.2-ostree-1.0 ? | 14:08 |
| juergbi | and python3-gi | 14:08 |
| juergbi | (debian package names) | 14:09 |
| tlater[m] | finn: You also need to install gobject-introspection | 14:10 |
| tlater[m] | Which you will need to install through distro packages | 14:10 |
| tlater[m] | As well as, well, ostree | 14:11 |
| tlater[m] | Neither of those things are available through pip | 14:11 |
| finn | got ostree, first thing I tried | 14:11 |
| tlater[m] | finn: If you *have* those things installed, your distro may not by default export ostree's gobject-introspection bindings, in which case you'd need to install something along the lines of ostree-dev | 14:12 |
| finn | If these are core deps, shouldn't they be in here? https://buildstream.build/source_install.html#installing_dependencies | 14:12 |
| gitlab-br-bot | marge-bot123 merged MR !1367 (bschubert/cython-disable-linetrace->master: Setup.py: Disable linetrace by default and only enable when using coverage) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1367 | 14:15 |
| tlater[m] | jennis: Are you sure ubuntu provides ostree dev bindings by default? | 14:16 |
| finn | maybe not the dev bindings | 14:16 |
| finn | hang on | 14:16 |
| tlater[m] | (I know you use ubuntu, so...) | 14:16 |
| tlater[m] | Doesn't look like it: https://packages.ubuntu.com/cosmic/libostree-dev | 14:16 |
| tlater[m] | finn: Try installing libostree-dev as well | 14:17 |
| *** lachlan has joined #buildstream | 14:18 | |
| finn | libostree-dev was what I needed | 14:19 |
| finn | thanks :) | 14:19 |
| finn | perhaps that doc should be updated | 14:19 |
| tlater[m] | finn: They are all optional dependencies because ostree is optional | 14:24 |
| finn | Oh sorry I thought it was part of bst 1.2? | 14:25 |
| finn | <jennis> Oh actually, you're using bst 1.2, that's in core | 14:25 |
| tlater[m] | But yes, it should be documented. Hopefully documentation for this will improve with the dependency split. | 14:25 |
| * tlater[m] really wonders what this chat log looks like on finn's end x) | 14:29 | |
| jennis | tlater[m], it used to be documented | 14:30 |
| finn | tlater[m], what do you mean? | 14:30 |
| tlater[m] | It does look like my client has a serious delay compared to the actual channel. | 14:30 |
| jennis | What I think happened was: OStree moved from core and this was taken out of the docs, and then the docs were split into 1.2 and master... so both versions don't have any mention of installing ostree | 14:31 |
| tlater[m] | finn: It's *part of* bst 1.2, but not everybody needs all plugins 1.2 provides. Hence plugin dependencies are not considered "core" | 14:31 |
| tlater[m] | It would be a shame if you needed to have gobject-introspection installed to run buildstream, when it can funcion just fine without. | 14:31 |
| finn | I see your point | 14:32 |
| finn | Thanks for the help anyway | 14:32 |
| jennis | but tlater[m], the 1.2 docs still have a "install plugin dependencies" section | 14:32 |
| jennis | and ostree deps are not present | 14:32 |
| jennis | https://buildstream.build/source_install.html#installing_dependencies | 14:33 |
| jennis | ostree deps used to be there too | 14:33 |
| *** xjuan has joined #buildstream | 14:34 | |
| tlater[m] | jennis: close. BuildStream used to *require* ostree (for artifact management), but when we changed that it no longer required ostree as a core dep | 14:35 |
| tlater[m] | I suspect whoever changed that doc forgot to update the plugin dependencies section - we'll definitely want to update that :) | 14:35 |
| jennis | tlater[m], those docs I just linked include, for each distro: "system requirements for the default plugins", for example they say for Debian you should `apt install bzr git lzip patch python3-arpy` | 14:47 |
| jennis | From what I remember, OSTree used to be included here, and the ostree dev libraries, but not anymore, so it seems | 14:47 |
| jennis | And way before that, yeah it was a core dependency | 14:47 |
| tlater[m] | jennis: I agree it's a bug in any case, hehe. Do you want to set up the ticket or should I? | 14:48 |
| jennis | I think we just need to revert a commit somewhere | 14:49 |
| tlater[m] | Well, unless someone wants to do that right now, we should have an issue telling us that | 14:49 |
| jennis | tlater[m], yeah I'll do it now | 14:52 |
| tlater[m] | jennis: I have a feeling this might be because the website is out of sync with the project... | 14:54 |
| tlater[m] | Looking into how these things are put together again | 14:54 |
| *** xjuan has quit IRC | 15:04 | |
| jennis | tlater[m]: https://gitlab.com/BuildStream/website/merge_requests/122 | 15:08 |
| tlater[m] | jennis: alas, I lack the authority to approve that :| | 15:11 |
| jennis | dang | 15:11 |
| jennis | tpollard ^^ ? | 15:12 |
| *** xjuan has joined #buildstream | 15:12 | |
| jennis | Seems you have power | 15:12 |
| tpollard | sec | 15:12 |
| tpollard | looks like you have a merge commit in that repo tlater[m] can you merge it? | 15:22 |
| tlater[m] | tpollard: I have a merge commit in that repo? | 15:23 |
| tlater[m] | Only a merge request | 15:23 |
| tlater[m] | In any case, that particular commit isn't that important atm, I'll merge it whenever someone approves. | 15:24 |
| tpollard | I've approved it :) but apparently I can't merge lol | 15:25 |
| tlater[m] | tpollard: I think you meant jennis ;p | 15:28 |
| *** lachlan has quit IRC | 15:28 | |
| * tpollard is confused, but it is friday afternoon | 15:29 | |
| *** lachlan has joined #buildstream | 15:31 | |
| tlater[m] | tpollard: It's friday rapidly-approaching-evening ;) | 15:31 |
| tpollard | afaict, your gitlab account has merged to the website repo | 15:32 |
| tpollard | so I don't understand why it doesn't allow you to approve xD | 15:32 |
| tlater[m] | Ahh, I see what you mean | 15:44 |
| tlater[m] | Yeah, no idea, I expect that the permissions are managed manually on that repo | 15:44 |
| tlater[m] | Not that I really mind, fewer things I have to look at ;p | 15:45 |
| tpollard | ;) | 15:47 |
| *** bochecha has quit IRC | 15:59 | |
| *** tpollard has quit IRC | 16:02 | |
| gitlab-br-bot | raoul.hidalgocharman opened issue #1041 (Support seperate end points for artifact caches) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1041 | 16:05 |
| *** raoul_ is now known as raoul | 16:08 | |
| *** finn has quit IRC | 16:26 | |
| *** finn has joined #buildstream | 16:29 | |
| *** phildawson has joined #buildstream | 16:30 | |
| *** xjuan has quit IRC | 16:33 | |
| *** raoul has quit IRC | 17:01 | |
| *** lachlan has quit IRC | 17:02 | |
| *** lachlan has joined #buildstream | 17:11 | |
| *** jonathanmaw has quit IRC | 17:38 | |
| *** slaf has quit IRC | 18:01 | |
| *** lachlan has quit IRC | 18:16 | |
| *** toscalix has quit IRC | 18:28 | |
| *** lachlan has joined #buildstream | 18:37 | |
| *** lachlan has quit IRC | 18:47 | |
| *** lachlan has joined #buildstream | 19:46 | |
| *** lachlan has quit IRC | 19:56 | |
| *** lachlan has joined #buildstream | 19:56 | |
| *** lachlan has quit IRC | 19:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!