IRC logs for #baserock for Friday, 2018-07-20

*** inara` has joined #baserock00:06
*** inara has quit IRC00:06
*** jjardon1 has quit IRC00:10
*** jjardon[m] has joined #baserock00:48
*** gtristan has joined #baserock04:43
*** gtristan has quit IRC07:34
gitlab-br-botlorries: merge request (franred/update-elfutils-repository->master: elfutils.lorry: Update repository) #37 changed state ("opened"):
gitlab-br-botlorries: merge request (franred/update-elfutils-repository->master: elfutils.lorry: Update repository) #37 changed state ("opened"):
gitlab-br-botlorries: merge request (franred/update-elfutils-repository->master: elfutils.lorry: Update repository) #37 changed state ("opened"):
gitlab-br-botlorries: merge request (franred/update-elfutils-repository->master: elfutils.lorry: Update repository) #37 changed state ("merged"):
noisecellmaybe we should think to add an script in the lorries repository which can be run by the CI and point to the repositories which are not longer reachable09:02
benbrown_seems a good idea, create an issue in the lorries repo09:03
* benbrown_ wonders if the http lorries should be changed to https09:04
* paulsherwood wonders if benbrown_ might understand/know the answer to
paulsherwoodActually, i think I do09:06
paulsherwoodand i think it's a bug... do you agree benbrown_09:06
gitlab-br-botlorries: issue #4 ("Check automatically if a repository is not longer reachable or used") changed state ("opened")
jjardonflatmush: I tried to work in that direction in the past, build as much as possible from git. Don't think you can replace any of the tarball with some additional work; I guess the next step is to bootstrap autotools, then rebuild on top of that09:08
flatmushyeah, that's what I expected, just make the bootstrap a bit bigger09:08
noisecellbenbrown_, done.09:09
benbrown_paulsherwood: Oh, yeah. Was discussing it with Tasos the other day. Ideally the .trees as found in definitions would override the one in the ybd base path.09:10
ironfootnoisecell: g.b.o is already publishing the mirroring failures09:13
noisecellironfoot, oh, where is this?09:14
benbrown_ironfoot: but g.b.o doesn't notify anyone, you'd have to check yourself09:15
benbrown_with gitlab ci you could setup a schedule, that'd email the owner of that schedule09:15
benbrown_on failure09:16
* noisecell would prefer to show that in an URL for each time we run it, instead of an email. so it is visible to everyone09:17
*** gtristan has joined #baserock09:19
benbrown_noisecell: you could always look to the schedules page, that'd show whether the most recent failed or not09:20
noisecellbenbrown_, what is that? :)09:20
noisecelloh, yeah, that makes sense09:22
*** jjardon[m] has quit IRC09:42
ironfootbenbrown_: we would get thousands of emails then...09:43
benbrown_ironfoot: AFAIA, only the owner of the pipeline would.09:44
benbrown_Which is a tad rubbish.09:44
benbrown_You'd only get a failure for the pipeline, not for every repo that might be broken.09:45
SotKusing email to report CI results is an antipattern which ends in all CI result emails being ignored ime09:46
*** jjardon[m] has joined #baserock09:52
ironfooti think the usecase wanted here is to be able to know if something is being mirrored correctly or not, in an easy manner: I think the UI exposed here helps with that:
ironfootnoisecell: ^09:53
benbrown_How often do any of you check that? :)09:55
benbrown_SotK: It would only be emailing on failure, but sure.09:55
noisecellironfoot, I see, that is useful, I might missed when reading the documentation09:56
noisecellmaybe worth to add to the README in lorries09:56
noisecelladd it*09:56
ironfootbenbrown_: is not a matter of keeping them maintained. Is only a matter of given the consumer of definitions some help to figure out if they are being updated09:57
noisecellironfoot, it is a combination of the url you posted and -- which shows the activity of the repository09:58
noisecellso you can discover that some repositories are still available but development has moved to a different location10:00
noisecell+ tarballs control10:00
benbrown_ironfoot: Why would we not want to keep them maintained?10:00
ironfootWell, users of g.b.o would want to keep it maintained, surely, and they can whenever they want10:01
ironfootmy point is: Is a lot of effort10:01
ironfootI've tried quite a few times in the past10:01
noisecellironfoot, I agree with you. I can see your point, I think we still thinking in trove as a part of definitions but we could make it extensive to other project, like right now we are using trove to get the provenance of our packages for trustable. Maybe this project will revive the interests on having the repositories lorried from the right places and updated10:17
ironfootthat would be great10:18
benbrown_Using http seems bad as far as trustability goes10:18
noisecellbenbrown_, I agree with yo10:18
*** inara` has quit IRC12:19
*** inara has joined #baserock12:32
*** gtristan has quit IRC13:18
*** noisecell has quit IRC14:22
*** rdale_ct has quit IRC15:59
*** toscalix has joined #baserock17:03
*** toscalix has quit IRC17:08
*** toscalix has joined #baserock17:13
*** toscalix has quit IRC18:34
*** jjardon[m] has quit IRC21:17
*** jjardon[m] has joined #baserock22:08

Generated by 2.15.3 by Marius Gedminas - find it at!