| *** jennis has joined #buildstream | 00:15 | |
| *** jennis_ has joined #buildstream | 00:15 | |
| *** jennis_ has quit IRC | 02:22 | |
| *** jennis has quit IRC | 02:22 | |
| *** ernestask has joined #buildstream | 05:02 | |
| *** noisecell has joined #buildstream | 05:07 | |
| *** tristan has joined #buildstream | 05:22 | |
| *** dominic has joined #buildstream | 06:52 | |
| *** noisecell has quit IRC | 07:36 | |
| *** toscalix has joined #buildstream | 07:50 | |
| *** toscalix_ has joined #buildstream | 07:55 | |
| *** toscalix has quit IRC | 07:55 | |
| *** toscalix_ has quit IRC | 08:03 | |
| *** toscalix_ has joined #buildstream | 08:05 | |
| *** finn has joined #buildstream | 08:08 | |
| *** bochecha_ has joined #buildstream | 08:24 | |
| *** finn_ has joined #buildstream | 08:28 | |
| *** finn has quit IRC | 08:30 | |
| *** toscalix has joined #buildstream | 08:31 | |
| *** toscalix_ has quit IRC | 08:33 | |
| *** jonathanmaw has joined #buildstream | 08:40 | |
| *** awacheux has joined #buildstream | 08:44 | |
| *** noisecell has joined #buildstream | 08:50 | |
| *** bethw has joined #buildstream | 08:51 | |
| juergbi | tlater: is there any test covering the artifact cache diff feature? | 09:07 |
|---|---|---|
| tlater | juergbi: I'm fairly sure there is, though I don't quite recall the name of the test... Hang. | 09:08 |
| juergbi | I thought there was as well but can't find it | 09:09 |
| tlater | juergbi: They are in integration/workspace.py | 09:09 |
| tlater | The last few tests | 09:09 |
| juergbi | hm, right, they should cover this, wondering why this doesn't work | 09:10 |
| tlater | What problem are you looking at? Always a chance I overlooked an edge case... | 09:12 |
| juergbi | never mind, I wasn't running the right tests | 09:12 |
| juergbi | it doesn't trigger added/removed subdirectories, afaict, but otherwise looks fine | 09:13 |
| tlater | Oh, right, I wanted to add a test case for that. | 09:13 |
| tlater | Damnit | 09:13 |
| gitlab-br-bot | buildstream: merge request (juerg/googlecas->master: WIP: Google CAS-based artifact cache) #337 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/337 | 09:14 |
| juergbi | jmac: the updated !337 should now pass all tests (except a CI docs issue I have to look into) | 09:16 |
| juergbi | it's still WIP, not ready for review, but would likely be usable as base for your branch, if you need it | 09:17 |
| jmac | Sounds great juergbi, I'll take a look | 09:17 |
| *** tiago has joined #buildstream | 09:25 | |
| jmac | juergbi: !445 might benefit from a quick look over from you to see if it's going in the right direction, if you have a chance | 09:25 |
| juergbi | indeed, planning to take a look today | 09:26 |
| *** toscalix has quit IRC | 09:46 | |
| gitlab-br-bot | buildstream: merge request (juerg/googlecas->master: WIP: Google CAS-based artifact cache) #337 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/337 | 09:48 |
| *** aday has joined #buildstream | 09:55 | |
| *** cs_shadow has joined #buildstream | 09:57 | |
| *** toscalix has joined #buildstream | 10:13 | |
| gitlab-br-bot | buildstream: merge request (214-filter-workspacing->master: Make workspace commands on a filter element transparently pass through to their build-dependency) #317 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/317 | 10:31 |
| juergbi | is there a difference in Python module search path or similar in Debian 8 / Python 3.4? | 10:41 |
| juergbi | pylint complains about missing in-tree module in CI, works fine on Debian 9 and Fedora 27 | 10:41 |
| cs_shadow | Hi. So, I have a couple of minor MRs that are awaiting review and "approval" in https://gitlab.com/BuildStream/buildstream-docker-images repository. Would anyone like to review them or should I bypass the approval process? | 10:44 |
| cs_shadow | tristan, jjardon: ^ | 10:44 |
| jonathanmaw | juergbi: can you review https://gitlab.com/BuildStream/buildstream/merge_requests/317/ when you have time? | 10:46 |
| juergbi | sure, adding it to mylist | 10:46 |
| tristan | cs_shadow, you can go ahead and merge what you think is appropriate and correct | 10:48 |
| tristan | cs_shadow, I know there is an "approval" feature that we dont use in the buildstream repo, you can use it in buildstream-docker-images if you like it, that is up to you :) | 10:48 |
| tristan | if you wanna disable it, please just disable it in the settings | 10:48 |
| *** jonathanmaw has quit IRC | 10:49 | |
| cs_shadow | tristan: thanks! I'll land these and will disable the approval feature for now. In case anyone wishes to introduce it again, we can surely discuss it then. | 10:50 |
| *** tristan has quit IRC | 11:01 | |
| *** jonathanmaw has joined #buildstream | 11:06 | |
| *** bethw has quit IRC | 11:34 | |
| *** tristan has joined #buildstream | 11:48 | |
| *** tristan_ has joined #buildstream | 11:48 | |
| *** toscalix has quit IRC | 12:03 | |
| *** jennis has joined #buildstream | 12:31 | |
| *** jennis_ has joined #buildstream | 12:31 | |
| *** bethw has joined #buildstream | 12:44 | |
| gitlab-br-bot | buildstream: merge request (kill_element_normal_name->master: Removed element.normal_name. (#288)) #465 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/465 | 13:19 |
| *** xjuan has joined #buildstream | 13:20 | |
| gitlab-br-bot | buildstream: merge request (kill_element_normal_name->master: WIP: Removed element.normal_name. (#288)) #465 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/465 | 13:41 |
| tristan_ | hmmm, I keep meaning to say something about that | 13:45 |
| finn_ | What would you call a name without an extension? At the moment self.name == "x.bst" | 13:55 |
| finn_ | in element.py | 13:55 |
| finn_ | Would you call x a name_prefix? | 13:57 |
| tristan_ | finn_, honestly, if we're going to replace normal_name with something else, it's not worth changing anything at all | 14:01 |
| tristan_ | I felt like calling a halt to this crusade against normal_name while it was assigned to milloni... but didnt get around to it | 14:02 |
| tristan_ | I think I was maybe overdoing it there | 14:02 |
| tristan_ | (with the crusade against normal_name) | 14:02 |
| finn_ | Agreed | 14:02 |
| valentind | tristan_, I suppose that tracking element cross junctions should always be done in project.refs. | 14:03 |
| valentind | I am trying to write tests for cross junction names for "track". | 14:03 |
| tristan_ | valentind, yes, and it currently does, I think there are regression tests for that too | 14:03 |
| finn_ | You end up having "x.bst" running throughout the code and end up a) removing the extension or b) creating a new "element.name_prefix" variable | 14:04 |
| tristan_ | valentind, there are early exits which ensure that the project uses project.refs if tracking was requested to cross junction boundaries | 14:04 |
| finn_ | tristan_, shall I close the issue and abandon? | 14:04 |
| tristan_ | valentind, for explicit `bst track junction.bst:element.bst` calls, I think those early checks need to be rethought, or ensured to still work | 14:04 |
| tristan_ | finn_, I agree to close and abandon | 14:05 |
| gitlab-br-bot | buildstream: issue #288 ("Kill Element.normal_name variable") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/288 | 14:05 |
| gitlab-br-bot | buildstream: merge request (kill_element_normal_name->master: WIP: Removed element.normal_name. (#288)) #465 changed state ("closed"): https://gitlab.com/BuildStream/buildstream/merge_requests/465 | 14:05 |
| tristan_ | valentind, also, I think that if a cross junction element was explicitly specified, I think that should probably... imply `-J` ? | 14:06 |
| valentind | Yes, I see. | 14:07 |
| tristan_ | valentind, i.e. we already specified a cross junction element, should it be implied that we intend to cross junction boundaries and the user should *not* have to specify -J ? | 14:07 |
| tristan_ | I think so | 14:07 |
| tristan_ | But... | 14:07 |
| tristan_ | What if we are doing multiple targets... --deps all | 14:07 |
| tristan_ | and one of them is not cross junction | 14:07 |
| tristan_ | I think the buck should still stop in recursion unless -J was specified | 14:07 |
| tristan_ | i.e. that is "Crossing the junction boundary" | 14:07 |
| tristan_ | not "specifying a cross junction element specifically" | 14:07 |
| *** toscalix has joined #buildstream | 14:07 | |
| valentind | I am not sure is actually needed. It is only for dependencies. | 14:08 |
| valentind | R | 14:08 |
| valentind | Right? | 14:08 |
| tristan_ | Right, so I think that means we agree | 14:08 |
| tristan_ | I.e. crossing junction boundaries is only about recursion | 14:08 |
| tristan_ | valentind, however, that probably also means sub-sub project recursion now | 14:08 |
| valentind | At least it looks like working without -J | 14:09 |
| valentind | No, actually it fails. | 14:09 |
| tristan_ | i.e. you can specify to recursively track junction.bst:element.bst | 14:09 |
| valentind | OK, so I will fix that. | 14:09 |
| tristan_ | now without -J, you probably dont want to recurse into *that* junction.bst project dependencies | 14:09 |
| tristan_ | there are a small hand full of edge cases to think about there and define policy for | 14:09 |
| *** laurence is now known as ltu | 14:20 | |
| *** finn_ has quit IRC | 14:22 | |
| *** finn has joined #buildstream | 14:22 | |
| *** noisecell has quit IRC | 14:56 | |
| *** tristan_ has quit IRC | 15:03 | |
| *** bethw has quit IRC | 15:08 | |
| *** bethw has joined #buildstream | 15:09 | |
| *** noisecell has joined #buildstream | 15:09 | |
| *** noisecell has quit IRC | 15:19 | |
| *** toscalix has quit IRC | 15:37 | |
| *** jennis_ has quit IRC | 15:53 | |
| *** jennis has quit IRC | 15:53 | |
| *** jennis has joined #buildstream | 15:58 | |
| *** jennis_ has joined #buildstream | 15:58 | |
| *** zilbermanka has quit IRC | 16:06 | |
| *** tristan_ has joined #buildstream | 16:12 | |
| *** jennis has quit IRC | 16:26 | |
| *** jennis_ has quit IRC | 16:26 | |
| *** toscalix has joined #buildstream | 16:28 | |
| *** bethw has quit IRC | 16:31 | |
| *** tristan_ is now known as tristan | 16:32 | |
| *** toscalix_ has joined #buildstream | 16:50 | |
| *** toscalix has quit IRC | 16:51 | |
| gitlab-br-bot | buildstream: merge request (valentindavid/359_cross_junction_elements->master: Allow names for cross junction elements) #454 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/454 | 16:52 |
| valentind | tristan, I added tests for pull, push, track, build, checkout, and show for cross junction elements. | 16:57 |
| valentind | Are there other commands I should add tests for? | 16:57 |
| finn | I'm a little confused by this. I've got an element which looks like this: | 17:28 |
| finn | kind: autotools | 17:29 |
| finn | sources: | 17:29 |
| finn | - kind: tar | 17:29 |
| finn | ur: some_tar.tar.gz | 17:30 |
| finn | s/ur/url | 17:30 |
| finn | directory: /home | 17:30 |
| tristan | valentind, there is fetch :) | 17:30 |
| tristan | valentind, but sounds awesome \o/ :) | 17:31 |
| finn | Shouldn't autotools be then run within the /home directory? | 17:31 |
| tlater | finn: Use some paste-* start tool, if you can, makes it a bit easier to read | 17:31 |
| tristan | finn, nope | 17:31 |
| tristan | finn, directory on the source, actually should not be allowed to be an absolute path afaics | 17:32 |
| tristan | maybe that's a bug, should probably trigger an error | 17:32 |
| tristan | finn, dont try to control an absolute path where things should happen, but if you want autotools commands (or any build element commands) to happen in a subdir of the build area, use %{command-subdir} | 17:33 |
| tristan | since it's common to all build elements, it's documented here: http://buildstream.gitlab.io/buildstream/buildstream.buildelement.html#module-buildstream.buildelement | 17:34 |
| tristan | instead of redundantly in every build element subclass | 17:34 |
| finn | thanks | 17:34 |
| tristan | command subdir is useful for cases like, when you build mozjs, you want to run commands in the javascript engine subdir of the mozilla source tree (for example) | 17:35 |
| *** jonathanmaw has quit IRC | 17:48 | |
| *** toscalix_ has quit IRC | 18:16 | |
| *** slaf has quit IRC | 18:20 | |
| *** slaf has joined #buildstream | 18:23 | |
| *** slaf has joined #buildstream | 18:24 | |
| *** dominic has quit IRC | 18:38 | |
| *** slaf_ has joined #buildstream | 19:34 | |
| *** slaf_ has joined #buildstream | 19:34 | |
| *** slaf has quit IRC | 19:37 | |
| *** slaf_ is now known as slaf | 19:37 | |
| *** xjuan has quit IRC | 20:01 | |
| *** Prince781 has joined #buildstream | 20:01 | |
| *** Prince781 has quit IRC | 20:04 | |
| *** Prince781 has joined #buildstream | 20:06 | |
| *** aday has quit IRC | 20:06 | |
| *** ernestask has quit IRC | 20:07 | |
| *** tristan has quit IRC | 20:10 | |
| *** Prince781 has quit IRC | 20:13 | |
| *** bochecha_ has quit IRC | 22:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!