| *** tristan has quit IRC | 04:05 | |
| *** LaurenceUrhegyi has joined #buildstream | 06:43 | |
| *** tristan has joined #buildstream | 07:53 | |
| *** ChanServ sets mode: +o tristan | 07:55 | |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 1 commit (last: _sandboxbwrap.py: Mount /dev/random in the sandbox) https://gitlab.com/BuildStream/buildstream/commit/02ca80b5c46a5a1bf96873575d33dab8ceb8340b | 08:05 |
|---|---|---|
| *** tristan has quit IRC | 08:09 | |
| *** tristan has joined #buildstream | 08:15 | |
| *** jonathanmaw has joined #buildstream | 08:38 | |
| *** ssam2 has joined #buildstream | 08:45 | |
| *** jonathanmaw has quit IRC | 08:54 | |
| *** tiagogomes has joined #buildstream | 08:59 | |
| *** LaurenceUrhegyi has quit IRC | 09:03 | |
| *** jonathanmaw has joined #buildstream | 09:04 | |
| *** jonathanmaw_ has joined #buildstream | 09:05 | |
| jjardon[m] | Hi, Im trying again to build definitions with buildstream in gitlab, but Im getting a problem with ldconfig now: https://gitlab.com/baserock/definitions/builds/16834560 Have you seen that one before? | 09:26 |
| * jjardon[m] updates https://gitlab.com/baserock/definitions/issues/9 | 09:27 | |
| * tristan thinks this sounds familiar | 09:28 | |
| * tristan looks | 09:29 | |
| tristan | jjardon[m], that is a very strange error actually | 09:40 |
| * tristan has never seen that, we _did_ have some ldconfig trouble but that was a matter of paths | 09:40 | |
| tristan | in this case, it looks like you dont have permission to execute the shell ??? | 09:40 |
| *** LaurenceUrhegyi has joined #buildstream | 09:49 | |
| *** LaurenceUrhegyi has quit IRC | 10:33 | |
| *** LaurenceUrhegyi has joined #buildstream | 10:46 | |
| *** Chris has quit IRC | 10:48 | |
| *** Chris has joined #buildstream | 11:04 | |
| *** LaurenceUrhegyi has quit IRC | 11:30 | |
| *** LaurenceUrhegyi has joined #buildstream | 11:38 | |
| jonathanmaw_ | tristan: getting back to the bzr source, AIUI it wasn't working properly (https://gitlab.com/BuildStream/buildstream/issues/9), and I had needlessly confused things by overloading the track field, where I should have encoded the branch name in the ref field, in additional to a way of identifying the revision. | 12:20 |
| tristan | Hmmm | 12:31 |
| tristan | jonathanmaw_, not sure that is exactly it... search for "Definitions of url, ref and track" in https://gitlab.com/BuildStream/buildstream/issues/10 | 12:32 |
| tristan | still I should document this in some plugin author facing docs | 12:33 |
| tristan | probably on the Source docs | 12:33 |
| tristan | Note: | 12:33 |
| tristan | A Project can be fetched and built if no track was ever specified for one or more Sources | 12:33 |
| tristan | A project cannot be built if there is no ref specified for one or more Sources | 12:33 |
| tristan | A source will have a ref if it was successfully tracked | 12:33 |
| tristan | A source which lacks a track configuration, cannot be tracked, but may be built if there is already a ref | 12:33 |
| tristan | jonathanmaw_, so basically I think this: If you *must* have a branch name to complete the ref information, and the exact ref information can be automatically updated using that branch name... then it is OK to require the 'track' attribute | 12:36 |
| tristan | jonathanmaw_, makes sense ? | 12:36 |
| tristan | I.e. I dont know about bzr really, but if the complete reference to an exact source checkout is 'trunk + a.b.c' (where a.b.c is meaningless without specifying 'trunk'), then yes please just require that 'track' be specified | 12:38 |
| tristan | if a.b.c is enough to specify an exact version, then no need to require track be specified for the purposes of fetching and staging | 12:38 |
| tristan | the bigger problem with the current approach was that 'ref' needed to be specified in order to run `bst track` | 12:39 |
| jonathanmaw_ | tristan: ok, having done some testing, I am certain that you need to specify the branch to get a unique revision | 12:46 |
| jonathanmaw_ | unless we want to throw deduplication to the wind, and pull branches as stand-alone entities | 12:46 |
| jonathanmaw_ | but I imagine that would upset some people when they switch branches and find they have to pull a whole load of redundant code | 12:47 |
| tristan | Right I see | 12:50 |
| tristan | Sure, so that is really fine, still; the 'ref' should not be required | 12:50 |
| tristan | Not for Source.track() to work, anyway | 12:50 |
| jonathanmaw_ | tristan: ok, that definitely seems to be a problem. | 12:51 |
| tristan | jonathanmaw_, ummm, explain ? | 12:52 |
| tristan | or, it was a problem in the existing code | 12:52 |
| tristan | yes that was my main objection | 12:52 |
| jonathanmaw_ | tristan: that I'm requiring 'ref'. I can't think of any reason why I should need to. | 12:53 |
| tristan | right, agreed :) | 12:53 |
| tristan | it was very confusing to me too when I tried it out | 12:53 |
| jonathanmaw_ | also, for the sake of showing-my-working, my conclusion that branches need to be considered to find a unique revisions is: in a repo, I have the branch "foo", which has 4 revisions, adding files a,b,c and d. I branch that from revision 2, and add a new revision that adds the file "q". The revision adding "q" is revno 3. When I create a third branch (quux) from foo, and merge baz into it, the commit that added "q" has revno 2.1.1, and revno 5 w | 12:55 |
| jonathanmaw_ | as the merge commit for pulling it into that branch. | 12:55 |
| *** violeta has joined #buildstream | 13:15 | |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 1 commit (last: widget.py: Pretty message when abbreviating log details) https://gitlab.com/BuildStream/buildstream/commit/f82a9be8f50490afa972468b1dad9aa5a5f7d5fb | 13:42 |
| *** tristan has quit IRC | 13:47 | |
| *** LaurenceUrhegyi has quit IRC | 14:00 | |
| *** tristan has joined #buildstream | 14:09 | |
| *** ChanServ sets mode: +o tristan | 14:09 | |
| tristan | Ummm... https://gitlab.com/BuildStream/buildstream/issues/9 | 14:10 |
| tristan | " @tristanvb | 14:10 |
| tristan | added ~1806484 Enhancement labels 3 weeks ago " | 14:10 |
| tristan | Really, I did that many ? | 14:11 |
| ssam2 | approximately | 14:16 |
| * paulsher1ood was unsuccessful ih his submission for a guadec talk... | 14:17 | |
| * paulsher1ood wonders if tristan fared any better | 14:17 | |
| tristan | yep my talk is approved :) | 14:18 |
| paulsher1ood | cool! | 14:19 |
| tristan | ssam2, you may know this... the email from registration@guadec.org said I should "confirm" but did not say how they want me to confirm, I replied to the email directly | 14:19 |
| tristan | I guess that's what they meant, there was nothing else at the website | 14:20 |
| tiagogomes | At least some people are confirming it using registration@guadec.org | 14:24 |
| tristan | gitlab folks say that ~1806484 is a label id for a non-existent label (so it used to be "In Progress" or something like that as well) | 14:24 |
| tristan | I found that other label to be just basically useless so I just deleted that other label, and that is the result | 14:25 |
| tristan | maybe it was a green "Doing" | 14:25 |
| ssam2 | tristan, yeah just reply to the email | 14:25 |
| ssam2 | tis a bug in the email I guess | 14:26 |
| ssam2 | paulsher1ood, I was a bit surprised to be honest! the focus seems to be on talks about specific GNOME technologies, which is kinda always the case | 14:26 |
| tristan | meta-gnome-apps-tested.bst - building webkit, built 123 of 132 modules, this time with one failure (and I think this one is a legitimate one, would also fail with jhbuild) | 14:32 |
| paulsher1ood | ssam2: never mind :) | 14:32 |
| jonathanmaw_ | paulsher1ood: what was your talk going to be about? | 14:39 |
| paulsher1ood | jonathanmaw_: non-developer perspective (from codethink) on why we contribute to and sponsor things in gnome | 14:41 |
| tristan | would have been good I think; can help to dispel the popular fairy tale that contributed developer hours grow on trees; help to cultivate a safe venue for contributions, etc | 14:43 |
| *** LaurenceUrhegyi has joined #buildstream | 15:01 | |
| *** LaurenceUrhegyi has quit IRC | 15:06 | |
| *** LaurenceUrhegyi has joined #buildstream | 15:09 | |
| *** jonathanmaw_ has quit IRC | 15:30 | |
| *** jonathanmaw has quit IRC | 15:30 | |
| *** LaurenceUrhegyi has quit IRC | 15:48 | |
| *** LaurenceUrhegyi has joined #buildstream | 15:49 | |
| *** LaurenceUrhegyi has quit IRC | 15:57 | |
| *** LaurenceUrhegyi has joined #buildstream | 16:13 | |
| *** LaurenceUrhegyi has quit IRC | 16:31 | |
| *** LaurenceUrhegyi has joined #buildstream | 16:33 | |
| *** ssam2_ has joined #buildstream | 16:59 | |
| *** ssam2 has quit IRC | 16:59 | |
| *** LaurenceUrhegyi has quit IRC | 18:07 | |
| *** ssam2_ has quit IRC | 18:14 | |
| jjardon[m] | paulsher1ood: there are going to be slots for talks outside the official programme; normally the people put them in a board and people vote | 18:19 |
| *** gitlab-br-bot has quit IRC | 18:27 | |
| *** gitlab-br-bot has joined #buildstream | 18:27 | |
| *** tristan has quit IRC | 19:31 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!