| *** bochecha has quit IRC | 00:11 | |
| *** Prince781 has joined #buildstream | 00:37 | |
| *** Prince781 has quit IRC | 01:22 | |
| *** Freejack has joined #buildstream | 01:30 | |
| *** Roedy has joined #buildstream | 01:45 | |
| *** MetaNova has joined #buildstream | 01:55 | |
| *** rud0lf has joined #buildstream | 02:25 | |
| *** leopi has joined #buildstream | 04:19 | |
| *** tristan has joined #buildstream | 04:28 | |
| *** leopi has quit IRC | 05:53 | |
| tristan | hmmmm.... [00:53:56][Pull:bootstrap-junction.bst:bootstrap-platform-extract.bst] | 06:05 |
|---|---|---|
| tristan | almost one hour of pulling | 06:05 |
| tristan | juergbi, maybe we could get this to stream the objects in one request while pulling ? | 06:06 |
| tristan | I think the round trips are killing me | 06:07 |
| *** leopi has joined #buildstream | 06:18 | |
| gitlab-br-bot | buildstream: issue #554 ("Pulling artifacts is very slow with high latency connections") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/554 | 07:14 |
| tristan | juergbi, filed #554 above regarding slowness of pull | 07:15 |
| tristan | tiagogomes, Thanks a lot for fixing freedesktop-sdk ! | 07:17 |
| tristan | jaysus... 25 minutes downloading the git artifact... and counting... | 07:33 |
| tristan | jjardon, would you say that 554 is an accurate description ? do things perform well for you when you happen to be on the same continent as the artifact cache server ? | 07:34 |
| Kinnison | tristan: is there any provision in the TUI libraries for progress bars? | 07:35 |
| juergbi | tristan: yes, there is definitely room for improvement. right now the download code is a bit too simple/serial | 07:37 |
| tristan | juergbi, if it was just me, I'd say this is not really urgent; but I think this is a pretty huge impact to users not in the EU :-S | 07:39 |
| Kinnison | Could we usefully put something like cloudflare in front of the artifact cache now that it's CAS based? | 07:39 |
| tristan | Kinnison, for the progress bar stuff, we'll probably want something custom in _frontend/status.py | 07:40 |
| * Kinnison nods | 07:40 | |
| tristan | Kinnison, click does "progress bars", but only in the simplest sense | 07:40 |
| Kinnison | I was just thinking any progress indication is useful | 07:40 |
| tristan | like, hashmarks for a single progressing thing | 07:40 |
| Kinnison | Helps you see if you're stalled, or just slow | 07:41 |
| Kinnison | Yeah, you really want a bar, an n/m indicator, and a progress rate indicator | 07:41 |
| tristan | indeed, I did ask about this recently in #git, and was told that I should look into using the --porcelain option, for parsable output | 07:41 |
| Kinnison | [######### ] 15/103 (250KiB/s) | 07:41 |
| Kinnison | yeah, --porcelain will help for processing git output | 07:41 |
| Kinnison | Or using pygit2 | 07:41 |
| * Kinnison was a little surprised that buildstream doesn't use pygit2 | 07:42 | |
| tristan | I'm a bit picky about the UI not taking too much screen realestate to be honest | 07:42 |
| tristan | I usually have 4 fetchers and 4 builders, and it only takes a few lines of status | 07:42 |
| * Kinnison nods | 07:43 | |
| tristan | if we needed a whole line for each ongoing task, we wont see much of the rolling log :-S | 07:43 |
| tristan | However... there is --verbose/--no-verbose | 07:43 |
| Kinnison | I was thinking a single progress bar line for any active fetch | 07:43 |
| Kinnison | s/any/all/ | 07:43 |
| * Kinnison -> appointment, back in about 90m | 07:43 | |
| tristan | And the ostree plugin does do per-second STATUS messages which should only appear in verbose mode | 07:43 |
| juergbi | Kinnison: cloudflare, not sure whether that would work well with gRPC | 07:44 |
| *** toscalix has joined #buildstream | 07:46 | |
| *** tiagogomes has quit IRC | 07:51 | |
| *** tiagogomes has joined #buildstream | 07:55 | |
| gitlab-br-bot | buildstream: merge request (phil/add-ubuntu-ci-job->master: WIP: .gitlab-ci-yml: Add ubuntu 18.04 test) #523 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/523 | 08:03 |
| *** noisecell has joined #buildstream | 08:26 | |
| gitlab-br-bot | buildstream: merge request (adamjones/systemd-cas->master: Include systemd file examples for google-cas cache) #524 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/524 | 08:26 |
| *** leopi has quit IRC | 08:27 | |
| gitlab-br-bot | buildstream: merge request (willsalmon/versionTagRegrex->master: Search for tags with the *.*.* patten for version) #601 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/601 | 08:28 |
| *** WSalmon has quit IRC | 08:34 | |
| *** WSalmon has joined #buildstream | 08:34 | |
| gitlab-br-bot | buildstream: merge request (Qinusty/502->master: cascache.py: Raise ArtifactError on grpc error) #576 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/576 | 08:37 |
| *** leopi has joined #buildstream | 08:42 | |
| toscalix | so checking https://gitlab.com/groups/BuildStream/-/boards/580444?milestone_title=BuildStream_v1.2&= we have 8 issues we want to resolve for RC that has no taker yet | 08:54 |
| toscalix | plus what it comes this week from people using beta | 08:55 |
| *** tiagogomes has quit IRC | 08:56 | |
| tristan | toscalix, #537 and #538 should probably be looked at by Jonathan Maw | 08:56 |
| tristan | those are bugs in his recently landed feature branch | 08:57 |
| toscalix | I will assign them to him, thanks | 08:57 |
| toscalix | and ping him | 08:57 |
| *** jonathanmaw has joined #buildstream | 08:57 | |
| tristan | speak of the devil :) | 08:57 |
| jonathanmaw | good morning | 08:58 |
| jonathanmaw | what's up? | 08:58 |
| tristan | jonathanmaw, We were hoping you'd look into #537 and #538 | 08:58 |
| toscalix | hi jonathanmaw there are two bugs considered urgent...those two, we would like you to evaluate | 08:58 |
| tristan | we hope to have those fixed before 1.2 and people start trying to use source mirroring | 08:58 |
| jonathanmaw | okie doke | 08:59 |
| tristan | jonathanmaw, asides from that... (and no hurry), it would be nice to formalize the release process in bst-external | 08:59 |
| tristan | as you are the maintainer, it would be good to say, maintain a NEWS file there, and announce the releases when you make tags etc, with the standard release mail jargon | 09:00 |
| tristan | but anyway; let's keep that on our radar | 09:00 |
| *** bochecha has joined #buildstream | 09:00 | |
| tristan | looks like GNOME is building reasonably well... still downloading llvm and rust, but those artifacts are big and didnt block other builds (other builds are now pending on it) | 09:02 |
| *** rdale has joined #buildstream | 09:04 | |
| Kinnison | juergbi: aah, I thought retrieval from the cache was simple GET methods | 09:04 |
| *** rdale has quit IRC | 09:06 | |
| jonathanmaw | tristan: okay, looking at #537, it seems the most straight-forward way would be to pass an alias override as an arg for Source.get_source_fetchers() | 09:08 |
| jonathanmaw | I gather we haven't made any promises about the API for get_source_fetchers() yet, so I can just add the arg? | 09:09 |
| *** rdale has joined #buildstream | 09:11 | |
| tristan | eek | 09:14 |
| tristan | jonathanmaw, this requires an API change ? | 09:14 |
| tristan | jonathanmaw, technically; *today* is feature freeze | 09:14 |
| jonathanmaw | tristan: I can work around it if need be | 09:15 |
| tristan | So let me understand the why for a moment | 09:15 |
| tristan | it seems to me, the get_source_fetchers() provides objects to the core which will be given resolved urls based on their stated aliases (or based on the originally "marked" urls) | 09:16 |
| tristan | jonathanmaw, whatever we do; we should NOT design the API around how the git plugin happens to be originally designed | 09:16 |
| Nexus | please could i get a review on https://gitlab.com/BuildStream/buildstream/merge_requests/504 Relative workspaces | 09:16 |
| tristan | jonathanmaw, our API design decision should be guided by what makes best sense as an API, and git should be made to conform to that | 09:17 |
| jonathanmaw | tristan: okay. would it be fair to say that a source may need to download something to be able to return the list of source fetchers? | 09:18 |
| jonathanmaw | Currently, the git source needs to be able to fetch its current ref to know which submodules exist | 09:18 |
| jonathanmaw | so that it can populate the list of source fetchers | 09:18 |
| tristan | right, I understand that part | 09:19 |
| tristan | but then, it's not really true afaics | 09:19 |
| tristan | jonathanmaw, the git plugin has explicit configuration allowing you to specify aliased urls for the various submodules | 09:19 |
| tristan | So that part is certainly known | 09:19 |
| tristan | jonathanmaw, if the project did not alias the submodules with plugin configuration... only then do we not know | 09:20 |
| tristan | and in that case, source mirroring is futile anyway | 09:20 |
| jonathanmaw | tristan: okay | 09:20 |
| tristan | jonathanmaw, so I guess it does not really make sense to say that... because for any source plugin... the same would be true; you have to have aliased URLs in the config if you want to leverage mirroring | 09:21 |
| tristan | jonathanmaw, from what I can tell though; there is still important inference | 09:23 |
| tristan | jonathanmaw, i.e.; thinking about that; the main repo needs to know what refs it needs; which is not in the config, so the API contract of source features should probably be explicit about respecting the ordering (if it is not already) ? | 09:24 |
| tristan | i.e. we dont ever bother going ahead with the second fetcher in the list, if the first one could not get it's content from any of the mirrors, and so on through the list | 09:25 |
| *** tiagogomes has joined #buildstream | 09:25 | |
| valentind | tristan, I suppose "patch" source plugin can create files so non determinism should be also fixed there. | 09:35 |
| tristan | valentind, interesting point; I missed that | 09:36 |
| tristan | valentind, I guess generally a patch file does not dictate the mode bits of the files it creates, or might not ? | 09:36 |
| tristan | valentind, please file a new issue for it then, and mark it related to the other issues in that group | 09:37 |
| valentind | OK | 09:37 |
| tristan | bochecha, I missed that about the docs change; I'm ambivalent, would you like to merge that in master now and then update it later after releasing 1.2 and pushing the packages to the official repo ? | 09:38 |
| tristan | bochecha, if so... please feel free to go ahead :) | 09:39 |
| bochecha | tristan: I'm ambivalent as well | 09:39 |
| bochecha | on one hand, a copr is not really great for proper maintenance (no bug report facility, it "feels" unofficial, etc…) | 09:40 |
| bochecha | on the other hand, having it in the doc will avoid someone else wasting time on redoing the work without knowing it's already done | 09:40 |
| tristan | good point | 09:40 |
| gitlab-br-bot | buildstream: issue #555 ("patch source plugin can introduce non-determinism") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/555 | 09:40 |
| tristan | bochecha, I commented now on the issue, feel free to unmark WIP and merge that | 09:40 |
| tristan | 2hours of pulling llvm :-S | 09:42 |
| tristan | I can build llvm in ~25min | 09:43 |
| gitlab-br-bot | buildstream: merge request (Qinusty/563-cache-quota-restriction->bst-1.2: Backport cache quota restriction to 1.2) #607 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/607 | 09:47 |
| gitlab-br-bot | buildstream: merge request (Qinusty/563-cache-quota-restriction->bst-1.2: Backport cache quota restriction to 1.2) #607 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/607 | 09:49 |
| tristan | jonathanmaw, I cancelled your pipeline, trying to expedite qinusty's backport | 09:52 |
| gitlab-br-bot | buildstream: merge request (bochecha/fedora-packages->master: doc: Mention the Fedora packages) #604 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/604 | 09:52 |
| tristan | jonathanmaw, (your pipeline does not have any commits beyond master so, I presume there is nothing to test anyway) | 09:52 |
| qinusty | 1 more backport to come from my previously merged MR's. I'll get it ready for when that one finishes tristan | 09:52 |
| *** rdale has quit IRC | 09:53 | |
| gitlab-br-bot | buildstream: merge request (bochecha/fedora-packages->master: doc: Mention the Fedora packages) #604 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/604 | 09:53 |
| jonathanmaw | tristan: correct, it was automatically started because I told gitlab to create a branch for me | 09:53 |
| bochecha | tristan: I rebased and removed the WIP | 09:53 |
| *** rdale has joined #buildstream | 09:54 | |
| tristan | bochecha, you have merge power right ? | 09:54 |
| bochecha | I do? :o | 09:54 |
| bochecha | ah, yes :x | 09:54 |
| * bochecha didn't know | 09:54 | |
| tristan | bochecha, please click the "Remove source branch" and press the "merge when pipeline succeeds button" :) | 09:54 |
| bochecha | I always check the "remove source branch" before submitting a MR :) | 09:55 |
| bochecha | just did the "merge when pipeline" | 09:55 |
| tristan | great | 09:55 |
| tristan | qinusty, which is the other backport btw ? | 09:55 |
| qinusty | The indiciations. E.g. "Pushing <artifact> -> <remote>", pulling etc | 09:56 |
| qinusty | https://gitlab.com/BuildStream/buildstream/merge_requests/553 | 09:56 |
| tristan | ah nice | 09:56 |
| qinusty | Just throwing it together now, I'm gonna throw in a second merge into the one backport as it was an MR to fix a mistake in that MR | 09:56 |
| tristan | qinusty, I don't know it if will make it into 1.1.5 | 09:56 |
| qinusty | that's okay. I'll just get it backported so it's all good to go when ready | 09:57 |
| tristan | I have to wrap everything up by 9pm, push the right buttons and have a dinner appointment | 09:57 |
| tristan | but indeed, it's more important that it gets backported, than making it specifically into 1.1.5 :) | 09:57 |
| jjardon | jonathanmaw: Hey! would it be possible to have another tag of bst-external? We are about to release the stable branch of freedesktop-sdk and It would be nice to recommend our users a specific tag intead "Use what we use in the CI" | 09:58 |
| jonathanmaw | jjardon: I can make a new tag now, and do a release announcement later in the day, if that works for you | 09:59 |
| tristan | jjardon, I sort of started that conversation earlier; I was hoping that we could get a NEWS file and start sending the standard release mails... you know "What is bst-external"... "New in 0.3"... etc | 09:59 |
| tristan | anyway I don't really mind if you tag now and we start tracking releases after :D | 09:59 |
| jjardon | jonathanmaw: tristan that will work perfectly for us, thanks! | 09:59 |
| jjardon | no massive rush from our part, we have not branched yet (we will attemp to do it today / tomorrow, though) | 10:00 |
| tristan | jonathanmaw, imminently, I will try to document our release process which could help you draft your own... there are a few steps | 10:01 |
| jjardon | As I said, It's more for convenience for people trying to build the runtime themselves than anything else | 10:01 |
| tristan | like a git command which tells you all the contributors since the last tag, etc | 10:01 |
| tristan | but not today haha, today I better just get this ball rolling | 10:02 |
| jonathanmaw | coldtom: it looks like the only difference between 0.2 and 0.3 will be your branch, coldtom/flatpakimage | 10:02 |
| jonathanmaw | Can you give me a succinct description of what it does that I can add to NEWS? | 10:02 |
| tristan | jjardon, from a freedesktop-sdk consumer of bst-external perspective, I think I'd still recommend that you use bst-external as a git submodule; that would let you load the plugins as "local", and ensure that anyone building the project is using exactly the version of plugins you decided | 10:03 |
| jonathanmaw | coldtom, and can you explain to me what's happening in https://gitlab.com/BuildStream/bst-external/commit/457d99ac39e830bb6f56ac1905d4c8d22f8c62de ? it looks like something that oughtn't to be there, to me | 10:04 |
| jjardon | tristan: you are probably right, I will try to create a MR to go in that direction | 10:06 |
| coldtom | jonathanmaw: The branch makes the flatpak image plugin create the directories for each extension defined in the metadata when built | 10:07 |
| tristan | jjardon, in the future it should be possible to have plugins installed separately, but (A) bst-external not being stable prevents that... (B) I still think that the way we load plugins from pip installs of the packages is not bullet proof enough to make sane guarantees | 10:07 |
| tiagogomes | submodules are pain. Honestly I'd prefer that we recommended that we installed them through pip | 10:08 |
| tristan | funny we went through a round of hardening before 1.0, but still it's tough | 10:08 |
| coldtom | jonathanmaw: the commit before that changed the cache key so as to force rebuilds after the patch was applied, valentind recommended I simplify it into just a version number rather than adding the extensions into the key | 10:08 |
| coldtom | i probably should have squashed the commits in retrospect | 10:08 |
| tristan | tiagogomes, yeah, but see above... we really need to know that... plugin that you use do not bump their artifact version, or break format | 10:08 |
| jonathanmaw | coldtom: ah, I see | 10:09 |
| jonathanmaw | jjardon: tag pushed for 0.3 https://gitlab.com/BuildStream/bst-external/commits/0.3 | 10:16 |
| tristan | valentind, looks like we missed the NEWS entry for include directives | 10:17 |
| tristan | valentind, no worry, I will add an entry in bst-1.2 for now | 10:17 |
| gitlab-br-bot | buildstream: issue #556 ("Terminal autocompletion for buildstream") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/556 | 10:17 |
| tristan | qinusty, I feel bad that you have been living a hard life of bst without completions; we should fix the install guide so that it's more discoverable again | 10:23 |
| tristan | it should definitely be visible in the ToC of the install page | 10:23 |
| qinusty | I use fish personally, but good to hear there are some completions to work off of :P I'll be looking into configuring fish for it and share whatever I come up with | 10:27 |
| tristan | qinusty, the click folks refused my patch to enhance the completions, so we have some evil hooks in place to get a nice completion UX | 10:28 |
| qinusty | :( | 10:28 |
| gitlab-br-bot | buildstream: merge request (Qinusty/563-cache-quota-restriction->bst-1.2: Backport cache quota restriction to 1.2) #607 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/607 | 10:34 |
| tristan | qinusty, please don't forget to update #556 | 10:35 |
| qinusty | Doing as we speak :D | 10:36 |
| gitlab-br-bot | buildstream: merge request (Qinusty/491->master: Cache quota is now restricted to available disk space) #563 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/563 | 10:37 |
| * tristan complains out loud again: [02:53:36][5092a480][ pull:freedesktop-sdk.bst:desktop/llvm6.bst] SUCCESS freedesktop-sdk/desktop-llvm6/5092a480-pull.15013.log | 10:37 | |
| tristan | 3 hours to download an llvm build ?! | 10:37 |
| tristan | wow | 10:37 |
| *** cs_shadow has joined #buildstream | 10:46 | |
| gitlab-br-bot | buildstream: issue #557 ("Add fish autocompletion support") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/557 | 10:48 |
| qinusty | Can we fix the issue TASK template for checkboxes? They currently have two spaces in them which means they don't actually work as intended. | 10:48 |
| toscalix | qinusty: there is a MR for it | 10:49 |
| qinusty | Ah okay | 10:49 |
| toscalix | let me reassign it to somebody else | 10:49 |
| qinusty | an MR or an issue? | 10:49 |
| gitlab-br-bot | buildstream: issue #558 ("Add autocompletion support for zsh") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/558 | 10:49 |
| tristan | toscalix, I think I commented, you made a patch, I did not have time to babysit the pipelines for it | 10:49 |
| toscalix | tristan: ah, ok | 10:50 |
| tristan | toscalix, feel free to press the rebase button and merge it - that action might need to be repeated when you are racing against multiple accepted merge requests | 10:50 |
| tristan | (i.e.: that is the "babysitting" part) | 10:50 |
| toscalix | ok | 10:50 |
| gitlab-br-bot | buildstream: merge request (toscalix->master: removed space in task template so checkboxes work as expected) #588 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/588 | 10:51 |
| qinusty | Slightly irritating that Gitlab can't rebase until merged | 10:52 |
| qinusty | but I guess that could cause issues | 10:52 |
| tristan | qinusty, I havent tested this to my knowledge, but when you have two subsequent MRs, you might rebase (locally on the git CLI) the second one against the first one, and merge the second one at the same time if it's pipeline starts after | 10:53 |
| tristan | I don't know if gitlab will recognize it's automatic merge commit for the first one, and then auto-rebase and apply | 10:53 |
| qinusty | Hmmmm, as long as the first one finishes first | 10:53 |
| gitlab-br-bot | buildstream: merge request (bochecha/fedora-packages->master: doc: Mention the Fedora packages) #604 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/604 | 10:53 |
| qinusty | Actuall no, because it makes a merge commit | 10:54 |
| tristan | qinusty, yeah that's my theory :) | 10:54 |
| qinusty | So it'll need a rebase for the merge commit | 10:54 |
| tristan | I mean.. gitlab made that merge commit, so it should be able to know that it's meaningless and doesnt effect the git state | 10:54 |
| qinusty | You'd think | 10:54 |
| tristan | it should be possible for it to know that - I haven't tested how smart it is | 10:54 |
| tristan | it worked before, when we didnt have the merge commits enabled | 10:54 |
| gitlab-br-bot | buildstream: merge request (toscalix->master: removed space in task template so checkboxes work as expected) #588 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/588 | 10:56 |
| gitlab-br-bot | buildstream: merge request (Qinusty/553-backport-1.2->bst-1.2: Backport cascache push/pull messages) #608 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/608 | 11:09 |
| gitlab-br-bot | buildstream: merge request (Qinusty/553-backport-1.2->bst-1.2: Backport cascache push/pull messages) #608 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/608 | 11:09 |
| phildawson | Nexus, I've added a couple of comments to !504 :) | 11:22 |
| *** mohan43u has quit IRC | 11:37 | |
| *** toscalix_ has joined #buildstream | 11:42 | |
| *** toscalix has quit IRC | 11:42 | |
| valentind | tristan, should we fix utimes for determinism in source plugins? | 11:45 |
| *** mohan43u has joined #buildstream | 11:45 | |
| tristan | valentind, utimes ... hmmm, I doubt it; we do update mtimes in a single place, this should not be in source plugins | 11:46 |
| valentind | sorry I meant mtime. | 11:46 |
| tristan | valentind, so that is already taken care of | 11:47 |
| valentind | OK, great. | 11:47 |
| valentind | So it is only mode, and making sure not to copy extended attributes. | 11:47 |
| gitlab-br-bot | buildstream: merge request (toscalix->master: removed space in task template so checkboxes work as expected) #588 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/588 | 11:49 |
| cs_shadow | Hi, I was thinking of publishing BuildStream on PyPi (there was a ML thread about it a while back). I have a sample version at https://test.pypi.org/project/BuildStream. It shouldn't require more than a couple minor changes to setup.py. | 11:49 |
| cs_shadow | I will of course submit an MR for it but I was wondering which account should we use on real PyPi? Should I create a "buildstream" account or use my own? | 11:49 |
| cs_shadow | tristan: tlater: you may be interested ^ | 11:49 |
| * tlater thinks a buildstream account would make more sense | 11:50 | |
| tristan | cs_shadow, I guess a PyPi means of distribution, if it has any chance of working at all... will not be detrimental, even if it cannot really be a recommended means of distribution ? | 11:50 |
| tristan | there is a whole elaborate issue about that... | 11:51 |
| tristan | https://gitlab.com/BuildStream/buildstream/issues/332 | 11:51 |
| tristan | of course a buildstream account is more appropriate | 11:51 |
| cs_shadow | tristan: yeah, it is not to solve the distribution problem, to make it easier to pip install/upgrade the BuildStream package. Rest of the steps won't change | 11:52 |
| cs_shadow | do we have a shared mailbox of sorts for BuildStream or should I just put my own? (or anyone else's if they'd like to own it?) | 11:53 |
| persia | Any particular reason not to use the mailing list? | 11:53 |
| tlater | I'm guessing this is one of those password-resetty account emails | 11:54 |
| cs_shadow | sorry, I should have phrased that better. I meant for creating the buildstream account on pypi. I assume they'll want to send verification mails etc that may not be suitable for the mailing luist | 11:54 |
| cs_shadow | list* | 11:54 |
| *** tristan has quit IRC | 11:55 | |
| tlater | cs_shadow: I haven't actually used pypi much, is there a concept of "organizations"? | 11:55 |
| tlater | The best option would be a set of permissions you can give to people | 11:56 |
| cs_shadow | tlater: not that I'm aware of. The closest it gets is having a group of people as maintainers that has to be managed manually | 11:56 |
| cs_shadow | i see two options broadly speaking: | 11:58 |
| cs_shadow | 1. create a generic buildstream account where we share login details | 11:58 |
| cs_shadow | 2. separately manage a list of people as maintainers of the pypi package | 11:58 |
| cs_shadow | there can be a separate account for CI in either case but I don't have any strong preference other than that | 11:58 |
| tlater | cs_shadow: Seems like numpy does both: https://pypi.org/project/numpy/ | 11:58 |
| tlater | A shared account that created the project | 11:58 |
| tlater | And a list of maintainers that can maintain it. | 11:59 |
| tlater | Which brings us back to the original question - who gets emails for this? | 11:59 |
| cs_shadow | i think the maintainer field of the package (not the account) should be the ML | 12:00 |
| cs_shadow | which is what numpy (and others) do, like "Numpy Developers" | 12:01 |
| *** tristan has joined #buildstream | 12:01 | |
| tlater | cs_shadow: My only concern is that you (or anyone owning the shared email inbox) might disappear in a black hole, making it impossible to access the account. | 12:02 |
| tristan | cs_shadow, I don't know how useful that will be, but maybe if we can recommend that instead of direct git; then we'll have a clearer story about what tags we recommend to users | 12:02 |
| *** leopi has quit IRC | 12:03 | |
| cs_shadow | tristan: using this instead of git is exactly what I had in mind, without replacing any other steps. If we upload wheels, we also won't require users to have setuptools. | 12:04 |
| cs_shadow | tlater: That's the same concern I have. I am not sure how can we mitigate that. (Is it the same for the Docker Hub too?) | 12:04 |
| tristan | cs_shadow, lets discuss this tomorrow... I'm especially interested to know how gracefully the `pip install` fails when the host dependencies are not found | 12:05 |
| tristan | and how it instructs the user to obtain those | 12:05 |
| tlater | cs_shadow: docker hub has organizations, and we have 2 active admins currently (one of them being you ;)) | 12:05 |
| tlater | I think if the maintainer feature works the way I think it does, it should be fine for you to create the project and make a few trusted people maintainers. | 12:06 |
| cs_shadow | tlater: Oh right! I was thinking that buildstream was a user in Docker Hub, my bad. I think if we do this, making a select number of maintainers should work fine. | 12:07 |
| cs_shadow | tristan: Sure, I will try to ping you tomorrow | 12:07 |
| *** jonathanmaw_ has joined #buildstream | 12:09 | |
| *** jonathanmaw has quit IRC | 12:09 | |
| *** mohan43u has quit IRC | 12:15 | |
| *** jonathanmaw_ is now known as jonathanmaw | 12:15 | |
| gitlab-br-bot | buildstream: merge request (Qinusty/553-backport-1.2->bst-1.2: Backport cascache push/pull messages) #608 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/608 | 12:18 |
| *** mohan43u has joined #buildstream | 12:18 | |
| gitlab-br-bot | buildstream: merge request (Qinusty/275->master: Indicate where artifacts are going to and coming from in the log) #553 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/553 | 12:19 |
| *** mohan43u has quit IRC | 12:25 | |
| *** ChanServ sets mode: +o tristan | 12:25 | |
| *** tristan changes topic to "/BuildStream 1.1.5 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" | 12:25 | |
| *** mohan43u has joined #buildstream | 12:29 | |
| cs_shadow | On a separate topic, in case someone has time, it'll be great to get some reviews on https://gitlab.com/BuildStream/buildstream/merge_requests/568 as rebasing it currently a pain :) | 12:31 |
| tristan | 1.1.5 \o/ | 12:31 |
| tristan | time for dinner :D | 12:31 |
| jonathanmaw | well, that's step one, I've got a test case to verify whether git sources can fetch with a missing upstream. | 12:35 |
| jonathanmaw | which also highlights the ostree failure | 12:35 |
| Nexus | woo, interesting exception \o/ | 12:40 |
| Nexus | if you run `bst` in a dir that does not exist, it throws a nce little exception | 12:40 |
| *** leopi has joined #buildstream | 12:41 | |
| laurence | I am really feeling the pain of not being able to click that link to gitlab in their emails :( | 12:48 |
| laurence | you never realised how much you value something until it's gone... | 12:48 |
| bochecha | laurence: same here :( | 12:49 |
| bochecha | something recently? | 12:49 |
| toscalix_ | laurence: what happened? Which change led to that missing link? | 12:51 |
| gitlab-br-bot | buildstream: merge request (Qinusty/backport-skipped-562->bst-1.2: Backport SKIPPED MessageType and duplicate skipped message fix) #609 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/609 | 12:51 |
| cs_shadow | toscalix_: it's an upstream issue, see https://gitlab.com/gitlab-com/production/issues/389#note_92335714 for details | 12:52 |
| *** toscalix_ is now known as toscalix | 12:54 | |
| *** leopi has quit IRC | 12:55 | |
| toscalix | laurence: I use plain text and the link works fine, in case you need it | 12:55 |
| laurence | bochecha, toscalix yes, a bug with gitlab - filed upstream https://gitlab.com/gitlab-org/gitlab-ce/issues/49899 | 12:56 |
| laurence | hmm, that issue is closed though.... | 12:56 |
| laurence | woah, those guys have a lot of labels...! | 12:56 |
| cs_shadow | maybe it is pending deployment (?) | 12:58 |
| gitlab-br-bot | buildstream: merge request (jennis/fix_links_in_docs->master: Fix two links in the installing docs) #610 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/610 | 13:08 |
| gitlab-br-bot | buildstream: merge request (jennis/fix_links_in_docs->master: Fix two links in the installing docs) #610 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/610 | 13:10 |
| gitlab-br-bot | buildstream: merge request (jennis/fix_links_in_docs->master: Fix two links in the installing docs) #610 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/610 | 13:11 |
| jennis | I'm not sure if an MR is even required for such a trivial change, but better safe then sorry eh, could someone review ^^, please? | 13:16 |
| * jonathanmaw has a look | 13:17 | |
| jennis | ta jonathanmaw | 13:17 |
| jonathanmaw | looks good to me | 13:18 |
| * jonathanmaw hits the button | 13:18 | |
| jonathanmaw | Now we're *both* responsible for it! | 13:18 |
| jennis | :o well we've saved whoever was responsible for the broken links beforehand | 13:19 |
| *** xjuan has joined #buildstream | 13:21 | |
| qinusty | git blame | 13:25 |
| qinusty | s/blame/praise | 13:25 |
| jennis | hehe | 13:25 |
| phildawson | I think that was my fault :( | 13:26 |
| phildawson | s/I think// | 13:27 |
| phildawson | Thanks for fixing it jennis :) | 13:27 |
| jennis | aha nw | 13:28 |
| gitlab-br-bot | buildstream: merge request (jennis/fix_links_in_docs->master: Fix two links in the installing docs) #610 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/610 | 14:02 |
| *** gitlab-br-bot has quit IRC | 14:08 | |
| *** gitlab-br-bot has joined #buildstream | 14:08 | |
| *** xjuan has quit IRC | 14:16 | |
| tiagogomes | `bst workspace list` Zzzzzz | 14:20 |
| gitlab-br-bot | buildstream: merge request (willsalmon/CacheExpiryTest->master: Trying to mitigate a file system issue) #595 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/595 | 14:21 |
| *** gitlab-br-bot has quit IRC | 14:32 | |
| *** gitlab-br-bot has joined #buildstream | 14:32 | |
| *** xjuan has joined #buildstream | 14:34 | |
| *** leopi has joined #buildstream | 14:42 | |
| *** gitlab-br-bot has quit IRC | 14:51 | |
| *** gitlab-br-bot has joined #buildstream | 14:51 | |
| gitlab-br-bot | buildstream: merge request (Qinusty/backport-skipped-562->bst-1.2: Backport SKIPPED MessageType and duplicate skipped message fix) #609 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/609 | 14:57 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550->master: PullQueue: fix resource used) #611 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/611 | 15:03 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550->master: PullQueue: fix resource used) #611 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/611 | 15:03 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550-backport->master: PullQueue: fix resource used) #612 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/612 | 15:06 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550-backport->bst-1.2: PullQueue: fix resource used) #612 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/612 | 15:07 |
| qinusty | Gitlab really need to allow more customisation in their webhooks | 15:13 |
| *** gitlab-br-bot has quit IRC | 15:15 | |
| *** gitlab-br-bot has joined #buildstream | 15:15 | |
| * tiagogomes stops | 15:15 | |
| *** gitlab-br-bot has quit IRC | 15:18 | |
| *** gitlab-br-bot has joined #buildstream | 15:18 | |
| gitlab-br-bot | buildstream: merge request (relative_workspaces->master: WIP: Patch for issue #191 support relative workspaces) #504 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/504 | 15:18 |
| *** gitlab-br-bot has quit IRC | 15:20 | |
| *** gitlab-br-bot1 has joined #buildstream | 15:20 | |
| *** gitlab-br-bot1 has quit IRC | 15:30 | |
| *** gitlab-br-bot has joined #buildstream | 15:30 | |
| *** Cool_Fire has joined #buildstream | 15:40 | |
| *** xjuan_ has joined #buildstream | 15:47 | |
| *** xjuan has quit IRC | 15:49 | |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550-backport->bst-1.2: PullQueue: fix resource used) #612 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/612 | 15:57 |
| *** gitlab-br-bot has quit IRC | 15:59 | |
| *** gitlab-br-bot has joined #buildstream | 15:59 | |
| *** gitlab-br-bot has joined #buildstream | 16:00 | |
| *** gitlab-br-bot has joined #buildstream | 16:00 | |
| *** gitlab-br-bot has quit IRC | 16:03 | |
| *** gitlab-br-bot has joined #buildstream | 16:03 | |
| *** gitlab-br-bot has quit IRC | 16:06 | |
| *** gitlab-br-bot has joined #buildstream | 16:06 | |
| *** gitlab-br-bot has quit IRC | 16:07 | |
| *** gitlab-br-bot has joined #buildstream | 16:07 | |
| *** gitlab-br-bot has quit IRC | 16:07 | |
| *** gitlab-br-bot1 has joined #buildstream | 16:07 | |
| *** gitlab-br-bot1 has quit IRC | 16:09 | |
| *** gitlab-br-bot has joined #buildstream | 16:09 | |
| *** gitlab-br-bot has quit IRC | 16:14 | |
| *** gitlab-br-bot has joined #buildstream | 16:14 | |
| *** gitlab-br-bot has joined #buildstream | 16:14 | |
| *** gitlab-br-bot has joined #buildstream | 16:15 | |
| *** oknf[m] has quit IRC | 16:24 | |
| *** waltervargas[m] has quit IRC | 16:24 | |
| *** cgmcintyre[m] has quit IRC | 16:24 | |
| *** alatiera has quit IRC | 16:24 | |
| *** Demos[m] has quit IRC | 16:24 | |
| *** ssssam[m] has quit IRC | 16:24 | |
| *** awacheux[m] has quit IRC | 16:24 | |
| *** segfault3[m] has quit IRC | 16:24 | |
| *** mattiasb has quit IRC | 16:24 | |
| *** theawless[m] has quit IRC | 16:24 | |
| *** kailueke[m] has quit IRC | 16:24 | |
| *** rafaelff[m] has quit IRC | 16:24 | |
| *** connorshea[m] has quit IRC | 16:24 | |
| *** abderrahim[m] has quit IRC | 16:24 | |
| *** m_22[m] has quit IRC | 16:24 | |
| *** krichter[m] has quit IRC | 16:24 | |
| *** pro[m] has quit IRC | 16:24 | |
| *** asingh_[m] has quit IRC | 16:24 | |
| *** jjardon[m] has quit IRC | 16:24 | |
| *** inigomartinez has quit IRC | 16:24 | |
| *** albfan[m] has quit IRC | 16:24 | |
| *** ptomato[m] has quit IRC | 16:24 | |
| *** leopi has quit IRC | 16:25 | |
| *** toscalix has quit IRC | 16:26 | |
| *** abderrahim[m] has joined #buildstream | 16:32 | |
| *** gitlab-br-bot has quit IRC | 16:37 | |
| *** gitlab-br-bot has joined #buildstream | 16:37 | |
| *** gitlab-br-bot has quit IRC | 16:38 | |
| *** gitlab-br-bot has joined #buildstream | 16:38 | |
| *** gitlab-br-bot has quit IRC | 16:42 | |
| *** gitlab-br-bot has joined #buildstream | 16:42 | |
| *** gitlab-br-bot has joined #buildstream | 16:42 | |
| *** gitlab-br-bot has joined #buildstream | 16:42 | |
| gitlab-br-bot | buildstream: issue #550 ("pulling from the cache doesn't honor the fetchers setting") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/550 | 16:55 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-550->master: PullQueue: fix resource used) #611 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/611 | 16:55 |
| *** jennis has quit IRC | 16:55 | |
| *** j1mc-polari has left #buildstream | 17:04 | |
| *** gitlab-br-bot has quit IRC | 17:04 | |
| *** gitlab-br-bot has joined #buildstream | 17:04 | |
| *** xjuan_ is now known as xjuan | 17:11 | |
| *** WSalmon has quit IRC | 17:13 | |
| *** WSalmon has joined #buildstream | 17:13 | |
| jjardon | Hey jonathanmaw ; thanks for the new tag in bst-external repo, but I think you forgot to update the version at https://gitlab.com/BuildStream/bst-external/blob/master/setup.py#L33 | 17:15 |
| jonathanmaw | jjardon: good catch! ta! | 17:16 |
| jonathanmaw | I guess this means I need to create a 3.1 release :-/ | 17:16 |
| jonathanmaw | er, 0.3.1 | 17:17 |
| jjardon | jonathanmaw: yeah, I will go for that. Also, any chance we can include the bst-external dependencies in the buildstream-fedora docker image? I have to manually download stuff from pip in our CI at the moment: https://gitlab.com/freedesktop-sdk/freedesktop-sdk/-/jobs/87197521 | 17:17 |
| jjardon | I would go for that* | 17:18 |
| jonathanmaw | jjardon: new tag pushed https://gitlab.com/BuildStream/bst-external/commits/0.3.1 | 17:24 |
| jjardon | jonathanmaw: excellent, tvm | 17:25 |
| *** gitlab-br-bot has quit IRC | 17:26 | |
| *** gitlab-br-bot has joined #buildstream | 17:26 | |
| *** gitlab-br-bot has quit IRC | 17:35 | |
| *** gitlab-br-bot has joined #buildstream | 17:35 | |
| *** Turner92 has joined #buildstream | 17:55 | |
| Nexus | Can you use buildstream on ubuntu? | 18:06 |
| coldtom | yes | 18:07 |
| persia | Should be able to, yes. | 18:07 |
| persia | Depending on the Ubuntu release, you may need to update some things. | 18:08 |
| Nexus | kk, ty | 18:08 |
| coldtom | you may need to install ostree and bubblewrap from source | 18:08 |
| persia | Are new enough versions not in backports? | 18:08 |
| persia | Also, if using 1.1.5, does one still need ostree? | 18:09 |
| coldtom | I'm unsure, I'm going off the ubuntu install instructions in the docs | 18:10 |
| coldtom | it seems that they're not | 18:11 |
| persia | How new a bubblewrap is needed? 0.3.0 seems to be in the archive, based on http://archive.ubuntu.com/ubuntu/ubuntu/pool/universe/b/bubblewrap/ | 18:13 |
| coldtom | >0.1.2 apparently, so that should be sufficient | 18:14 |
| coldtom | s/>/>= | 18:14 |
| persia | Current LTS has 0.2.1, so yeah, that's new enough. | 18:16 |
| persia | So, if the CAS stuff means no more ostree, then it should just work | 18:17 |
| coldtom | how do you install that from the command line? | 18:17 |
| *** jonathanmaw has quit IRC | 18:38 | |
| jjardon | for bst >= 1.1.4, ostree is only needed if you use the ostree source plugin in your project | 18:44 |
| jjardon | bubblewrap is still a hard dependency: docs says 0.1.2 is enough (not sure that's a typo though): https://buildstream.gitlab.io/buildstream/install_linux_distro.html | 18:45 |
| jjardon | yes, seems to be correct: https://gitlab.com/BuildStream/buildstream/blob/master/setup.py#L45 | 18:49 |
| *** gitlab-br-bot has quit IRC | 18:55 | |
| *** gitlab-br-bot has joined #buildstream | 18:55 | |
| persia | Nexus: Summary: you should be able to use buildstream with just pip :) | 19:04 |
| *** ssssam[m] has joined #buildstream | 19:05 | |
| *** connorshea[m] has joined #buildstream | 19:05 | |
| *** cgmcintyre[m] has joined #buildstream | 19:05 | |
| *** albfan[m] has joined #buildstream | 19:05 | |
| *** inigomartinez has joined #buildstream | 19:05 | |
| *** segfault3[m] has joined #buildstream | 19:05 | |
| *** m_22[m] has joined #buildstream | 19:05 | |
| *** theawless[m] has joined #buildstream | 19:05 | |
| *** mattiasb has joined #buildstream | 19:05 | |
| *** alatiera has joined #buildstream | 19:05 | |
| *** awacheux[m] has joined #buildstream | 19:05 | |
| *** asingh_[m] has joined #buildstream | 19:05 | |
| Nexus | woo \o/ | 19:05 |
| *** waltervargas[m] has joined #buildstream | 19:05 | |
| *** pro[m] has joined #buildstream | 19:05 | |
| *** oknf[m] has joined #buildstream | 19:05 | |
| *** jjardon[m] has joined #buildstream | 19:05 | |
| *** Demos[m] has joined #buildstream | 19:05 | |
| *** kailueke[m] has joined #buildstream | 19:05 | |
| *** rafaelff[m] has joined #buildstream | 19:05 | |
| *** ptomato[m] has joined #buildstream | 19:05 | |
| *** krichter[m] has joined #buildstream | 19:05 | |
| Nexus | ty persia | 19:05 |
| persia | And matrix is back :) | 19:06 |
| *** gitlab-br-bot has quit IRC | 19:07 | |
| *** gitlab-br-bot has joined #buildstream | 19:07 | |
| *** gitlab-br-bot has quit IRC | 19:08 | |
| *** gitlab-br-bot1 has joined #buildstream | 19:08 | |
| *** gitlab-br-bot1 has quit IRC | 19:10 | |
| *** gitlab-br-bot has joined #buildstream | 19:10 | |
| *** gitlab-br-bot has quit IRC | 19:13 | |
| *** gitlab-br-bot has joined #buildstream | 19:13 | |
| *** gitlab-br-bot has joined #buildstream | 19:14 | |
| *** gitlab-br-bot has quit IRC | 19:14 | |
| *** gitlab-br-bot has joined #buildstream | 19:14 | |
| *** gitlab-br-bot has quit IRC | 19:15 | |
| *** gitlab-br-bot has joined #buildstream | 19:15 | |
| *** gitlab-br-bot has joined #buildstream | 19:15 | |
| gitlab-br-bot | buildstream: merge request (edbaunton/executable-remote-source->master: remote.py: Add support for marking downloaded files executable) #581 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/581 | 20:02 |
| *** tristan has quit IRC | 20:31 | |
| *** gitlab-br-bot has quit IRC | 20:47 | |
| *** gitlab-br-bot has joined #buildstream | 20:47 | |
| *** gitlab-br-bot has quit IRC | 20:50 | |
| *** gitlab-br-bot1 has joined #buildstream | 20:50 | |
| *** gitlab-br-bot1 has quit IRC | 21:14 | |
| *** gitlab-br-bot has joined #buildstream | 21:14 | |
| *** xjuan has quit IRC | 21:59 | |
| *** gitlab-br-bot has quit IRC | 22:51 | |
| *** gitlab-br-bot has joined #buildstream | 22:51 | |
| gitlab-br-bot | buildstream: merge request (jjardon/doc_arch_releases->master: source/install_linux_distro.rst: Make clearer ArchLinux packages available) #613 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/613 | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!