| gitlab-br-bot | buildstream: issue #88 ("Update the install doc with OSTree >= v2017.8") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/88 | 00:10 |
|---|---|---|
| gitlab-br-bot | buildstream: issue #88 ("Update the install doc with OSTree >= v2017.8") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/88 | 00:39 |
| gitlab-br-bot | push on buildstream@master (by Tristan Van Berkom): 2 commits (last: tests/frontend/buildcheckout.py: Run tests in strict/non-strict mode) https://gitlab.com/BuildStream/buildstream/commit/787fb42f962951756268afa5b6bebc0030e68d89 | 00:39 |
| *** tristan has quit IRC | 00:54 | |
| *** bochecha has quit IRC | 01:18 | |
| *** bochecha has joined #buildstream | 01:19 | |
| *** bochecha has quit IRC | 01:23 | |
| *** bochecha has joined #buildstream | 01:51 | |
| *** bochecha has quit IRC | 02:23 | |
| *** bochecha has joined #buildstream | 03:51 | |
| *** bochecha has quit IRC | 04:23 | |
| *** bochecha has joined #buildstream | 05:51 | |
| *** bochecha has quit IRC | 06:23 | |
| *** bochecha has joined #buildstream | 06:28 | |
| *** bochecha has quit IRC | 07:01 | |
| *** bochecha has joined #buildstream | 08:04 | |
| *** jonathanmaw has joined #buildstream | 08:43 | |
| *** tlater has joined #buildstream | 10:31 | |
| tlater | Heh, I can't reproduce the blessings failure when running the tests directly, without setup.py | 10:36 |
| tlater | This is a fun one | 10:37 |
| tlater | It's definitely the redirection through setup.py - presumably it closes something before running pytest. | 11:53 |
| tlater | ... No, maybe not | 11:54 |
| tlater | When forcing the tests to run through subprocess.call() it works fine. It's the way in which pytest is called by default. | 11:55 |
| tlater | I guess that makes sense since subprocess.call() would open a new stdin/stdout | 11:55 |
| tlater | Yup, this is how pytest-runner runs pytest: | 12:08 |
| tlater | with _save_argv(_sys.argv[:1] + self.addopts): | 12:08 |
| tlater | self.result_code = __import__('pytest').main() | 12:08 |
| *** tlater has quit IRC | 12:24 | |
| *** tlater has joined #buildstream | 13:10 | |
| *** tlater has quit IRC | 13:23 | |
| *** tlater has joined #buildstream | 13:23 | |
| *** xjuan has joined #buildstream | 13:26 | |
| *** xjuan has quit IRC | 13:35 | |
| *** xjuan has joined #buildstream | 13:38 | |
| tlater | Calling `sys.__stdout__ = open('/dev/stdout', 'w')` in the function that executes bst fixes the problem... Not sure if this hack is alright, but it's the best I've got. | 13:49 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: Fix tests for other platforms) https://gitlab.com/BuildStream/buildstream/commit/9ac88535b370ce7e69145e1417dfa2059ed2bd7c | 14:19 |
| gitlab-br-bot | buildstream: merge request (cross_platform->master: WIP: Cross platform) #81 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/81 | 14:19 |
| *** tlater has quit IRC | 14:22 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: Fix tests for other platforms) https://gitlab.com/BuildStream/buildstream/commit/d0169e205ac32d6ad618071979032e61fa029374 | 14:29 |
| gitlab-br-bot | buildstream: merge request (cross_platform->master: WIP: Cross platform) #81 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/81 | 14:29 |
| *** bochecha has quit IRC | 14:40 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: Fix tests for other platforms) https://gitlab.com/BuildStream/buildstream/commit/8fa354b40d41ecc480e4d412bd032aab7d8f646f | 14:42 |
| gitlab-br-bot | buildstream: merge request (cross_platform->master: WIP: Cross platform) #81 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/81 | 14:42 |
| *** tlater has joined #buildstream | 15:13 | |
| *** bochecha has joined #buildstream | 16:05 | |
| *** bochecha_ has joined #buildstream | 16:37 | |
| *** bochecha has quit IRC | 16:37 | |
| gitlab-br-bot | push on buildstream@validation (by Tristan Maat): 2 commits (last: Add _yaml.validate_node) https://gitlab.com/BuildStream/buildstream/commit/5948904371d31d09674271b3e5975188775168a7 | 16:51 |
| *** bochecha_ has quit IRC | 17:03 | |
| *** bochecha_ has joined #buildstream | 17:03 | |
| *** bochecha_ has quit IRC | 17:08 | |
| *** jonathanmaw has quit IRC | 17:09 | |
| *** bochecha_ has joined #buildstream | 17:32 | |
| *** tlater has quit IRC | 17:46 | |
| *** tristan has joined #buildstream | 18:11 | |
| *** ChanServ sets mode: +o tristan | 18:11 | |
| jjardon[m] | Hi, is there a way to tell buildstream to not push artifacts? Im trying to avoid https://gitlab.com/baserock/definitions/-/jobs/31550939 | 18:19 |
| jjardon[m] | https://buildstream.gitlab.io/ give a 404, is that expected? | 18:20 |
| gitlab-br-bot | buildstream: issue #89 ("Absolute build directories should be unique per element") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/89 | 18:22 |
| jjardon[m] | doesnt seem there is an option here: https://buildstream.gitlab.io/buildstream/config.html#config | 18:22 |
| tristan | jjardon[m], mmyeah currently I either configure it or not | 18:23 |
| tristan | jjardon[m], obviously if you're configuring the buildstream.conf already, you can comment out the push url | 18:24 |
| tristan | but there is nothing to do that on the CLI | 18:24 |
| tristan | jjardon[m], I am thinking of changing the configuration of artifact caches, though | 18:25 |
| tristan | So that there can be more than one, and I think I like the idea of having some kind of `bst remote add/remove` kind of api for it | 18:25 |
| jjardon[m] | Im currently trying to think of the best solution for https://gitlab.com/baserock/definitions/issues/13 | 18:27 |
| tristan | jjardon[m], I'm having a hard time to see what that issue is about | 18:27 |
| tristan | only a title | 18:27 |
| jjardon[m] | if you are not using a protected branch, the CI will fail. so normal contributors can not submit anything | 18:28 |
| tristan | Umm | 18:29 |
| jjardon[m] | it will fail because bst will return failure when trying to push the artifacts | 18:29 |
| jjardon[m] | as those branches doesnt have the key | 18:29 |
| tristan | So basically buildstream will fail to build if it is configured to push but cannot | 18:29 |
| tristan | is it ? | 18:29 |
| jjardon[m] | exactly | 18:29 |
| tristan | Hmmmyeah I see, currently it's coded to bail out | 18:31 |
| tristan | which is strange, because we have this litter of 'can_push()' all over the place | 18:31 |
| tristan | which I would think we wouldnt need if we were just going to bail out anyway | 18:31 |
| tristan | jjardon[m], well ok so a work around right now is to sed the .gitlab-ci.yml differently based on the branch right ? | 18:34 |
| tristan | jjardon[m], other than that I think it's a bug | 18:34 |
| tristan | but | 18:34 |
| jjardon[m] | maybe an option to "only warn instead fail" when push fail would work | 18:34 |
| jjardon[m] | tristan: yeah, but for that I have to duplicate all the jobs | 18:34 |
| jjardon[m] | one for "protected" branches, another for the rest | 18:34 |
| gitlab-br-bot | buildstream: issue #90 ("bst build bails if it cannot push to the artifact cache") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/90 | 18:36 |
| tristan | jjardon[m], really ? that sucks | 18:37 |
| tristan | jjardon[m], you cant have the branch as a variable and conditionally generate the buildstream.conf depending on the branch name ? | 18:37 |
| tristan | even without substing the .gitlab-ci.yml, just in the launch script | 18:38 |
| jjardon[m] | mmm, probably | 18:38 |
| jjardon[m] | time for dinner; I will try that later! thanks | 18:38 |
| tristan | anyway I filed a bug, I think by default we should just work | 18:38 |
| * jjardon[m] wants to spend some time with buildstream anyway :) | 18:38 | |
| *** xjuan has quit IRC | 18:41 | |
| tristan | jjardon[m], happen to still be around ? | 19:12 |
| tristan | I could use someone with gitlab talents | 19:12 |
| * tristan wants to cache the .coverage files and read them back in a later stage | 19:13 | |
| jjardon[m] | tristan: add them to artifacts: | 19:26 |
| jjardon[m] | AFAIK, all the artifacts are passed by default between stages | 19:26 |
| tristan | jjardon[m], can I read the artifact generated from a previous stage in the next ? | 19:26 |
| tristan | Where do I read it from ? | 19:26 |
| tristan | just expect it to be there ? | 19:27 |
| tristan | jjardon[m], I was taking an approach of dumping them in the cache directories | 19:27 |
| jjardon[m] | tristan: yes https://docs.gitlab.com/ee/ci/yaml/#artifacts | 19:27 |
| jjardon[m] | and yes they are passed by default: https://docs.gitlab.com/ee/ci/yaml/#dependencies | 19:28 |
| tristan | jjardon[m], freaking perfect | 19:28 |
| tristan | :) | 19:28 |
| *** bochecha_ has quit IRC | 21:40 | |
| *** tristan has quit IRC | 23:08 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!