| *** benschubert has quit IRC | 00:58 | |
| *** tristan has quit IRC | 05:55 | |
| *** benschubert has joined #buildstream | 06:37 | |
| *** tristan has joined #buildstream | 07:34 | |
| gitlab-br-bot | marge-bot123 merged MR !1488 (alexfazakas/fix-frontend-typo->master: app: Fix "earily" typo) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1488 | 07:58 |
|---|---|---|
| *** alexandrufazakas has joined #buildstream | 08:08 | |
| *** finn has joined #buildstream | 08:13 | |
| alexandrufazakas | Hey tristan, I'd appreciate it if you could have a look at my work on #178 whenever you can :) | 08:15 |
| alexandrufazakas | Or if anyone else takes a look and has some feedback, I'm all ears :D | 08:18 |
| Kinnison | Where is the work? | 08:19 |
| alexandrufazakas | Kinnison: This repository https://gitlab.com/AlexFazakas/learn-gitlab-pages | 08:19 |
| alexandrufazakas | Kinnison: and this is what it looks like https://alexfazakas.gitlab.io/learn-gitlab-pages/ | 08:20 |
| * Kinnison looks | 08:22 | |
| Kinnison | Is that just an index? | 08:23 |
| Kinnison | or should it be linking to versions of the docs? | 08:23 |
| alexandrufazakas | Kinnison: I don't understand the question, isn't there a list of the available versions? | 08:24 |
| alexandrufazakas | Oh okay | 08:24 |
| alexandrufazakas | It's broken apparently | 08:24 |
| alexandrufazakas | I broke something with my last push last night, apparenty | 08:24 |
| alexandrufazakas | s/apparenty/apparently | 08:25 |
| Kinnison | oops | 08:25 |
| Kinnison | :D | 08:25 |
| alexandrufazakas | Thanks for pointing it out :) | 08:25 |
| *** ChanServ sets mode: +o tristan | 08:27 | |
| tristan | alexandrufazakas, I was in the middle of that :) | 08:27 |
| tristan | alexandrufazakas, Comments up | 08:27 |
| alexandrufazakas | tristan: sweet, cheers | 08:27 |
| tristan | alexandrufazakas, looks like you figured a way to make it completely automated too, great :) | 08:27 |
| alexandrufazakas | tristan: There's a script in my repo which filters out the latest micro point of each release btw | 08:28 |
| tristan | In bash | 08:28 |
| tristan | yeah | 08:28 |
| alexandrufazakas | Yes | 08:28 |
| tristan | I think that is fine | 08:28 |
| alexandrufazakas | Is that a problem? | 08:28 |
| alexandrufazakas | I thought since I was already using bash for setting everything up, I might throw that in as well | 08:29 |
| tristan | It doesnt do the "only latest micro version for each minor point" thing | 08:29 |
| tristan | But in the end (as I said in my comments), maybe that is not so important | 08:29 |
| alexandrufazakas | tristan: I thought we wanted for each even minor point | 08:29 |
| tristan | minor point != micro point | 08:29 |
| benschubert | valentind: are you around? Anything I am blocking on the freedesktop side of things? (I'd like to be able to fix the overnight tests as soon as possible :) | 08:30 |
| alexandrufazakas | Okay | 08:30 |
| tristan | alexandrufazakas, Anyway check the comments, I think with the badges in there it will be pretty perfect, we can always refine things later | 08:30 |
| alexandrufazakas | tristan: Alright, I'll have a look at those | 08:30 |
| tristan | alexandrufazakas, for reference, see the badges at the top of the install page: https://buildstream.build/install.html | 08:32 |
| tristan | alexandrufazakas, the way that works fwiw... is that every time a pipeline runs on master, the static badges are updated to reflect the latest stable release and development snapshot - so any related docs material just embeds those badges and magically stay up to date :) | 08:33 |
| alexandrufazakas | tristan: That makes sense. Thanks a lot for having a look :) | 08:34 |
| alexandrufazakas | So these should be displayed on the page with all the versions, right? | 08:34 |
| alexandrufazakas | The badges, that is | 08:36 |
| tristan | alexandrufazakas, Yeah, since now it's just a huge list of versions, if a user ends up there they might be scratching their head wondering "Which should I pick ?" | 08:36 |
| tristan | alexandrufazakas, I think if they want "master", they will know, and if they want latest stable or latest dev, those are good choices to highlight | 08:36 |
| alexandrufazakas | tristan: I think that shouldn't be difficult to implement | 08:37 |
| tristan | alexandrufazakas, I also commented that it could be good to override the badge links in that case to point to the urls of their docs | 08:37 |
| tristan | Yeah should be dead easy :) | 08:37 |
| tristan | alexandrufazakas, The one part I didnt look at, is how the updates of the site get triggered | 08:38 |
| tristan | I guess a nightly job would be fine | 08:38 |
| alexandrufazakas | Yeah | 08:39 |
| alexandrufazakas | I think that would be a good idea | 08:39 |
| alexandrufazakas | Kinnison: FWIW, it should work as intended now :D | 08:40 |
| Kinnison | nice | 08:41 |
| *** tristan has quit IRC | 08:43 | |
| *** tristan has joined #buildstream | 08:43 | |
| alexandrufazakas | tristan: Should I just make all my changes to the docs-website repository and open a MR on it? | 08:45 |
| alexandrufazakas | It's not very clear to me how I'd see the changes in that case | 08:45 |
| alexandrufazakas | That's why I created my repository in the first place | 08:45 |
| *** ChanServ sets mode: +o tristan | 08:47 | |
| tristan | alexandrufazakas, Ok so I think (A) Your CI should have 2 steps... one of them which aggregates the docs and generates the index, and adds that to a "gitlab artifact", the other job depends on the first one and just publishes the artifact | 08:47 |
| tristan | alexandrufazakas, (B) The publish part which uses credentials to update the actual website at docs.buildstream.build should be only run for master branch | 08:47 |
| tristan | alexandrufazakas, Then... when you create the merge request, it will not update docs.buildstream.build, but we can browse the site in the resulting gitlab artifact in gitlab's pipeline UI | 08:48 |
| alexandrufazakas | tristan: I'm not sure what a gitlab artifact is, but I'll have a read on it. I think I get what you're saying | 08:49 |
| alexandrufazakas | I'll try to set up a MR in the meantime, then | 08:49 |
| finn | I've almost managed to execute Buildstream with Buildbarn (another RE server like BuildGrid) but it fails when collecting the output directories. It seems Buildbarn doesn't allow ".." as a directory path. Is there any reason Buildstream doesn't define the output directories relative to the input root instead? | 08:49 |
| tristan | alexandrufazakas, https://gitlab.com/BuildStream/buildstream/blob/master/.gitlab-ci.yml#L344 | 08:49 |
| tristan | alexandrufazakas, notice that the "publish" job depends on the "docs" job | 08:50 |
| alexandrufazakas | tristan: Ah, okay, yep | 08:50 |
| tristan | alexandrufazakas, and it's in a later "stage" | 08:50 |
| tristan | oh sorry, it's the "pages" job in the "publish" stage | 08:50 |
| tristan | the docs job should show you how to create the artifact | 08:50 |
| alexandrufazakas | Yep, I think it's clear | 08:51 |
| alexandrufazakas | And I'll use this to write my own | 08:51 |
| finn | I think I have my answer from the api: " // The paths are relative to the working directory of the action execution." | 08:56 |
| finn | So I'll raise an issue with Buildbarn and find out why relative paths aren't supported. Once they are, Buildstream should have a second compatible RE server :) | 08:57 |
| Kinnison | nice | 08:58 |
| finn | I can actually see the actions being successfully executed. It's just when it comes to uploading the output files =/ | 09:02 |
| *** jonathanmaw has joined #buildstream | 09:18 | |
| *** lachlan has joined #buildstream | 09:31 | |
| *** traveltissues_ has joined #buildstream | 09:31 | |
| *** traveltissues has quit IRC | 09:42 | |
| *** traveltissues has joined #buildstream | 09:43 | |
| valentind | benschubert, I am changing completely the docker images. I am hoping to get it to work today. Then we will have an image built with buildstream master. | 09:43 |
| benschubert | valentind: great, thanks a lot :) Ping me if there's a problem on the building BuildStream part :) | 09:43 |
| *** toscalix has joined #buildstream | 10:13 | |
| *** tristan has quit IRC | 10:16 | |
| *** lachlan has quit IRC | 10:19 | |
| Kinnison | WSalmon: I'm intending to go through !951 soon. Are you intending to rebase it and/or make changes for the pipeline fails in the next hour? | 10:23 |
| gitlab-br-bot | MR !951: Add BuildBox backend for sandboxing https://gitlab.com/BuildStream/buildstream/merge_requests/951 | 10:24 |
| WSalmon | i have just tweaked it and waiting on CI | 10:24 |
| Kinnison | I mean, it's red and static afaict, I wonder if my computer isn't updating | 10:25 |
| WSalmon | i am happy to leave it be for the next hour to allow review | 10:25 |
| Kinnison | It's also 44 commits behind master | 10:25 |
| WSalmon | oh no | 10:25 |
| WSalmon | it my other branch i have been playing with | 10:25 |
| WSalmon | on sec | 10:25 |
| Kinnison | I won't be able to complete a review before lunch, so I figure I'll let you fiddle and I'll come to it after lunch | 10:25 |
| WSalmon | sounds good | 10:26 |
| *** toscalix has quit IRC | 10:28 | |
| *** tristan has joined #buildstream | 10:30 | |
| *** lachlan has joined #buildstream | 10:39 | |
| *** phoenix has joined #buildstream | 10:40 | |
| WSalmon | benschubert, i am a bit baffled by https://gitlab.com/BuildStream/buildstream/-/jobs/255152012 i think you are our best pylint guru could you enlighten me as to what is wrong with it it seems to be conplaining about comments but there arnt any... i tried adding some extra but that didnt seem to help... not sure if you will have time but many thanks if you do | 10:41 |
| Kinnison | WSalmon: I see comments: https://gitlab.com/BuildStream/buildstream/blob/willsalmon/buildbox/src/buildstream/storage/_casbaseddirectory.py#L466 | 10:42 |
| *** traveltissues has quit IRC | 10:43 | |
| benschubert | Yep, you have 3 lines commented out | 10:43 |
| *** traveltissues_ has quit IRC | 10:45 | |
| *** traveltissues has joined #buildstream | 10:45 | |
| traveltissues | i'll also take a look WSalmon | 10:45 |
| *** lachlan has quit IRC | 10:57 | |
| alexandrufazakas | tristan: I don't think we can just wrap <a href> around a badge like that, it still points to the tarball | 11:01 |
| *** lachlan has joined #buildstream | 11:24 | |
| WSalmon | I think i had got in a fuss and my editor and linter looking at sutuly diffrent files, thanks Kinnison benschubert | 11:29 |
| lchlan | Hi, is it possible for anyone on channel to create a sub-project in gitlab for the best-benchmarking-bot? | 11:35 |
| *** juanalday has joined #buildstream | 11:35 | |
| * paulsherwood assumes lchlan means bst-benchmarking-bot | 11:38 | |
| lchlan | Apologies autouncorrect error. | 11:39 |
| *** finn has quit IRC | 12:02 | |
| valentind | lchlan, I think I can do that. | 12:20 |
| valentind | Should it be called bst-benchmarking-bot? | 12:21 |
| valentind | lchlan, And is your username on gitlab lchlan? | 12:23 |
| valentind | lchlan, I created that https://gitlab.com/BuildStream/bst-benchmarking-bot | 12:26 |
| valentind | and added you as maintainer for that project. Is that what you wanted? | 12:26 |
| *** phil has joined #buildstream | 12:30 | |
| *** phildawson_ has quit IRC | 12:30 | |
| lchlan | valentind: Thanks, that should be spot on. | 12:45 |
| *** finn has joined #buildstream | 12:53 | |
| alexandrufazakas | Can someone give me Developer access so I can open a branch on BuildStream/docs-website, please? | 12:54 |
| alexandrufazakas | tristan, juerbi ^ if either of you can | 12:55 |
| valentind | alexandrufazakas, What is your username on gitlab? | 12:56 |
| alexandrufazakas | valentind: AlexFazakas | 12:56 |
| valentind | alexandrufazakas, done. | 12:56 |
| alexandrufazakas | valentind: Thanks a lot! :) | 12:56 |
| Kinnison | WSalmon: Is your MR ready for me to go through now? | 13:20 |
| *** phoenix has quit IRC | 13:30 | |
| *** juanalday has quit IRC | 13:34 | |
| *** lachlan has quit IRC | 13:41 | |
| Kinnison | WSalmon: My brain has melted so I've not been through all the test changes yet | 13:43 |
| Kinnison | WSalmon: but there's a bunch'o'comments | 13:44 |
| WSalmon | thanks | 13:44 |
| WSalmon | much appreciated | 13:44 |
| WSalmon | also thank traveltissues | 13:44 |
| *** lachlan has joined #buildstream | 13:51 | |
| *** juanalday has joined #buildstream | 13:58 | |
| alexandrufazakas | tristan: I've opened my MR on docs-website. Didn't manage to make the badges work the way we wanted them to :( | 14:00 |
| alexandrufazakas | I'll see if I can group versions a bit better | 14:01 |
| *** traveltissues has quit IRC | 14:04 | |
| *** lachlan has quit IRC | 14:17 | |
| *** lachlan has joined #buildstream | 14:35 | |
| lchlan | benschubert: Hi I'm having a few issues with benchmarking and I'm told that you are working on a patch set to resolve. The errors are shown in https://gitlab.com/BuildStream/benchmarks/-/jobs/253492966. | 14:55 |
| benschubert | lchlan: hey, seems like some files are not recognized as having changed. doing a 'git clean -fdx '*.c' '*.so' would fix the problem | 14:57 |
| lchlan | benschubert: Seems to be a new issue introduced recently, but I can incorporate that in the script to see if is resolves the point. Thanks | 14:58 |
| *** traveltissues has joined #buildstream | 14:58 | |
| benschubert | lchlan: the problem is that cython is doing like make, basing themselves on timestamps | 14:59 |
| benschubert | (at least that's what I think where the problem comes from) | 14:59 |
| benschubert | I'll see if I manage to get a fix | 14:59 |
| benschubert | anyways, having a clean repo every time is better :) | 14:59 |
| lchlan | The tests are done in a clean docker instances on each occasion, but I will see if it has any effect | 15:02 |
| benschubert | ouch if that's the case, we have a problem | 15:04 |
| benschubert | I thought we have a single docker for all commits? | 15:05 |
| lchlan | One docker container per SHA for the nightly run rather than the weekly run by jennis. | 15:07 |
| benschubert | ok that's really weird :/ | 15:08 |
| lchlan | I will just check something | 15:13 |
| *** phildawson_ has joined #buildstream | 15:20 | |
| *** finn has quit IRC | 15:21 | |
| *** phil has quit IRC | 15:22 | |
| *** bochecha has quit IRC | 15:50 | |
| *** juanalday has quit IRC | 15:53 | |
| *** alexandrufazakas has quit IRC | 15:56 | |
| *** phildawson_ has quit IRC | 16:12 | |
| *** jonathanmaw has quit IRC | 16:48 | |
| *** lachlan has quit IRC | 16:55 | |
| lchlan | benschubert : looks like this is an issue with bubble wrap and the docker configuration. | 16:55 |
| benschubert | weird, nothing there has changed in quite a while | 16:56 |
| *** lachlan has joined #buildstream | 17:01 | |
| *** lachlan has quit IRC | 17:04 | |
| gitlab-br-bot | traveltissues opened issue #1087 (Reduce cache_keys_unstable manipulation) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1087 | 17:05 |
| *** lachlan has joined #buildstream | 17:12 | |
| *** tiagogomes has quit IRC | 17:39 | |
| gitlab-br-bot | marge-bot123 closed issue #1083 (source checkout fails when parent directory is not writable) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1083 | 17:52 |
| gitlab-br-bot | marge-bot123 merged MR !1491 (chandan/fix-source-checkout-1083->master: stream.py: Don't checkout sources in the parent of the location) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1491 | 17:52 |
| *** lachlan has quit IRC | 17:59 | |
| *** lachlan has joined #buildstream | 18:01 | |
| gitlab-br-bot | traveltissues opened (was WIP) MR !1492 (traveltissues/setstability->master: Add _set_cache_keys_stability method) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1492 | 18:03 |
| *** traveltissues has quit IRC | 18:20 | |
| gitlab-br-bot | marge-bot123 closed issue #1086 (bst workspace open: `--force` and `--no-checkout` don't work well together) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1086 | 18:29 |
| gitlab-br-bot | marge-bot123 merged MR !1490 (chandan/workspace-force-no-checkout->master: Don't remove workspace directory when `--no-checkout` option is given) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1490 | 18:29 |
| *** juanalday has joined #buildstream | 18:51 | |
| *** juanalday has quit IRC | 18:54 | |
| *** lachlan has quit IRC | 19:21 | |
| *** lachlan has joined #buildstream | 19:33 | |
| *** lachlan has quit IRC | 20:01 | |
| lchlan | benschubert: All very odd, reverted a number of elements and the issue bwrap persists. | 20:05 |
| *** slaf has quit IRC | 21:26 | |
| *** slaf has joined #buildstream | 21:29 | |
| *** slaf has joined #buildstream | 21:29 | |
| *** cs-shadow has quit IRC | 23:11 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!