| *** hasebastian has joined #buildstream | 03:40 | |
| *** narispo has quit IRC | 05:21 | |
| *** narispo has joined #buildstream | 05:21 | |
| *** narispo has quit IRC | 05:23 | |
| *** narispo has joined #buildstream | 05:24 | |
| *** hasebastian has quit IRC | 06:08 | |
| gitlab-br-bot | juergbi opened MR !1876 (juerg/centos-7.7->master: .gitlab-ci.yml: Update to CentOS 7.7) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1876 | 06:11 |
|---|---|---|
| *** tristan has quit IRC | 07:19 | |
| *** benschubert has joined #buildstream | 07:21 | |
| *** ikerperez has quit IRC | 07:25 | |
| *** cphang has quit IRC | 07:25 | |
| *** tristan has joined #buildstream | 07:32 | |
| *** ChanServ sets mode: +o tristan | 07:34 | |
| *** ikerperez has joined #buildstream | 07:38 | |
| *** cphang has joined #buildstream | 07:51 | |
| *** phildawson-ct has joined #buildstream | 08:06 | |
| tristan | juergbi, got a moment ? I'm trying to respond to your email about the junction jungle | 08:07 |
| juergbi | sure | 08:07 |
| tristan | juergbi, specifically I'm having a hard time parsing the part where you would reverse it... in my example I had foo's declaration of baz.bst junction explicitly overriding the bar project's junction baz.bst | 08:09 |
| juergbi | yes, so I'd rather specify in foo's bar.bst that its baz.bst should be overridden by foo's baz.bst | 08:10 |
| tristan | are you saying that you would prefer to have foo's bar.bst say "override this project's junction baz.bst... with this project local declaration: baz.bst" ? | 08:10 |
| juergbi | correct | 08:10 |
| tristan | Ok | 08:10 |
| tristan | :) | 08:10 |
| juergbi | as we're effectively modifying the bar subproject | 08:11 |
| *** santi has joined #buildstream | 08:17 | |
| *** tpollard has joined #buildstream | 08:17 | |
| *** lachlan has joined #buildstream | 08:19 | |
| *** rdale has joined #buildstream | 08:33 | |
| *** phoenix has joined #buildstream | 08:33 | |
| tristan | juergbi, got it out... maybe if we can polish this off I can work on it next after these versioning things | 08:47 |
| *** hasebastian has joined #buildstream | 08:54 | |
| *** phoenix has quit IRC | 08:59 | |
| *** phoenix has joined #buildstream | 09:05 | |
| *** hasebastian has quit IRC | 09:12 | |
| *** phoenix has quit IRC | 09:20 | |
| juergbi | tristan: replied. that would be great | 09:36 |
| *** cs-shadow has joined #buildstream | 09:45 | |
| *** tristan has quit IRC | 10:17 | |
| gitlab-br-bot | juergbi merged MR !1876 (juerg/centos-7.7->master: .gitlab-ci.yml: Update to CentOS 7.7) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1876 | 10:30 |
| *** tristan has joined #buildstream | 10:52 | |
| *** tristan has quit IRC | 11:03 | |
| *** phoenix has joined #buildstream | 11:05 | |
| *** hasebastian has joined #buildstream | 11:14 | |
| gitlab-br-bot | traveltissues opened issue #1289 (automatic cachekey updating for tests is broken) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1289 | 12:39 |
| WSalmon | juergbi, are you able to run tox locally for bst-plugins-experimental? im trying to tidy up git_tag and i still cant figure out whats up | 14:42 |
| gitlab-br-bot | juergbi closed issue #1289 (automatic cachekey updating for tests is broken) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1289 | 14:46 |
| *** lachlan has quit IRC | 14:52 | |
| *** lachlan has joined #buildstream | 14:52 | |
| *** phoenix has quit IRC | 14:53 | |
| juergbi | WSalmon: no, not working properly here for some reason, haven't investigated yet | 14:53 |
| WSalmon | ok, i have just got tpollard to run them and he has no issue, im really confused, i have tried to run them loads of diffrent versiond of python and tox and i just keep drawing a blank | 14:54 |
| juergbi | maybe there is an issue if bst-plugins-experimental is installed in the current venv | 14:55 |
| tpollard | if you want anything else just ask | 14:55 |
| WSalmon | thanks tpollard | 14:57 |
| WSalmon | i made a new venv and only installed tox | 14:57 |
| juergbi | WSalmon: and that works? | 14:58 |
| WSalmon | nope | 14:58 |
| WSalmon | everything i have tried ends up with bst-plugins-experimental==0.12.0,BuildStream==1.93.1 | 14:59 |
| juergbi | same here | 14:59 |
| WSalmon | i even tried removing .tox | 14:59 |
| WSalmon | im on fedora-31 is tpolard on ubuntu? | 14:59 |
| tpollard | 19.10 yeh | 15:00 |
| WSalmon | ooooooooooooo | 15:02 |
| WSalmon | i did a fresh clone and it works | 15:02 |
| WSalmon | juergbi, maybe a git clean with the right letters is needed.. | 15:03 |
| *** lachlan has quit IRC | 15:05 | |
| WSalmon | `git clean -fX` dosent seem to be enough tho | 15:06 |
| juergbi | WSalmon: I had a very old bst_plugins_experimental.egg-info/ | 15:06 |
| juergbi | git clean -dfX should suffice | 15:06 |
| WSalmon | im confused how that happened tho | 15:08 |
| WSalmon | yep `git clean -dfX` did it | 15:09 |
| WSalmon | benschubert, your our expert at tox do you have any idea how this could have happened? i dont mind adding a patch to try and stop it but i dont know how to start | 15:10 |
| jjardon | hi, I need to do a full rebuild of a bst project: what would be the best way? ie, best/cleanest way to force a change in the cache key | 15:33 |
| coldtom | jjardon open a workspace on the bottom element would be a way i guess? or configure buildstream to not look in your cache | 15:35 |
| coldtom | (that is, change the location of your cache to somewhere else | 15:36 |
| jjardon | ah, but I want the cache to be confgured and used, but not to pull basically | 15:36 |
| coldtom | ah, so you want to force a full build to be pushed to a remote cache? | 15:37 |
| cphang | can an environment variable be injected via project.conf? | 15:39 |
| coldtom | ^that should change cache keys aiui | 15:40 |
| juergbi | jjardon: is this for test/validation, or what's the reason for this? | 15:42 |
| tpollard | if you don't want to allow pull, set the fetchers to zero or --remote to something dummy depending on your bst version | 15:42 |
| jjardon | juergbi: test | 15:42 |
| jjardon | coldtom: exactly that :) | 15:43 |
| jjardon | tpollard: does that work with bst 1.4.2 ? | 15:43 |
| coldtom | tpollard, fetchers are also used for `fetch` aren't they? | 15:43 |
| coldtom | i guess you could do it in multiple steps | 15:44 |
| tpollard | coldtom: maybe, not 100% sure | 15:44 |
| tpollard | I would assume yes, that it's source fetchers too | 15:46 |
| coldtom | if it's possible to do the build first, then configure the cache, then push the artifacts i think that'd be the way to go, but then you will only push (--deps all) on successful builds (i guess you could ignore build failures and wait for the push failure when it tries to push non-existent artifacts) | 15:46 |
| *** hasebastian has quit IRC | 15:52 | |
| *** lachlan has joined #buildstream | 15:59 | |
| WSalmon | https://gitlab.com/BuildStream/bst-external/-/blob/master/bst_external/sources/git_tag.py#L564 this allows my plugin to know if the user selected false or didnt spesifiy false, if the user didnt specify false/true and the repo has gitlfs then i can raise a configureable warning. how do i do this with master, what do i call on node? i dont seem to be able to have a boolin default to None | 16:03 |
| *** lachlan has quit IRC | 16:05 | |
| WSalmon | its wiered having a boolin default to None but it works so well as i can still just do `if var:` and var can be None or false and it still behaves right but if i need to check if it has been spesified then i can expelicity check for None | 16:05 |
| juergbi | valentind: I can now build all fd-sdk oci images with the juerg/vdirectory branches of buildstream and bst-plugins-experimental | 16:10 |
| tpollard | WSalmon: maybe look at the ref not in track configurable warning? | 16:10 |
| juergbi | valentind: will you build them yourself for testing or shall I send you the artifacts/tarballs somehow? | 16:10 |
| juergbi | ideally, we would be testing them built with the default bwrap sandbox as well as BST_FORCE_SANDBOX=buildbox-run | 16:11 |
| juergbi | they shouldn't differ but exercise different code paths in buildstream core | 16:11 |
| tpollard | WSalmon: or the configurable submodule warnings | 16:11 |
| juergbi | WSalmon: I've recently added support for `None` as default value for ints | 16:18 |
| juergbi | we could likely do the same for bool | 16:18 |
| juergbi | alternatively, I think you could check something like: "use-lfs" in node | 16:19 |
| juergbi | and then make the get_bool call conditional | 16:19 |
| *** lachlan has joined #buildstream | 16:20 | |
| *** tristan has joined #buildstream | 16:27 | |
| *** douglaswinship has quit IRC | 16:27 | |
| benschubert | you also have `node.is_none()` which would work | 16:27 |
| benschubert | WSalmon: for the tox question, not sure what there is to fix? I don't know how this old bst-plugins-experimental version got installed there | 16:28 |
| gitlab-br-bot | juergbi closed MR !1874 (juerg/validate-path->master: _casbaseddirectory.py: Validate path components) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1874 | 16:30 |
| *** lachlan has quit IRC | 16:34 | |
| valentind | juergbi, I will test tomorrow. Thank you. | 16:47 |
| juergbi | ta | 16:47 |
| *** lachlan has joined #buildstream | 16:53 | |
| *** lachlan has quit IRC | 16:56 | |
| *** lachlan has joined #buildstream | 16:56 | |
| *** tpollard has quit IRC | 17:01 | |
| *** lachlan has quit IRC | 17:02 | |
| *** santi has quit IRC | 17:15 | |
| *** narispo has quit IRC | 17:35 | |
| *** phoenix has joined #buildstream | 17:51 | |
| *** cphang has quit IRC | 17:52 | |
| *** ikerperez has quit IRC | 17:52 | |
| *** bethw has quit IRC | 17:54 | |
| *** jward has quit IRC | 17:55 | |
| *** WSalmon has quit IRC | 17:55 | |
| *** valentind has quit IRC | 17:55 | |
| *** robjh has quit IRC | 17:55 | |
| *** jward has joined #buildstream | 17:56 | |
| *** robjh has joined #buildstream | 17:56 | |
| *** WSalmon has joined #buildstream | 17:56 | |
| *** valentind has joined #buildstream | 17:56 | |
| *** cphang has joined #buildstream | 17:59 | |
| *** bethw has joined #buildstream | 18:00 | |
| *** rdale has quit IRC | 18:09 | |
| *** ikerperez has joined #buildstream | 18:15 | |
| *** ceibal has joined #buildstream | 18:16 | |
| *** ceibal has quit IRC | 18:45 | |
| *** benschubert has quit IRC | 19:19 | |
| *** rdale has joined #buildstream | 19:25 | |
| *** phildawson-ct has quit IRC | 20:09 | |
| gitlab-br-bot | cs-shadow closed issue #1278 (Remote execution test is failing) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1278 | 20:10 |
| gitlab-br-bot | cs-shadow opened issue #1290 (Some tests break with pytest 5.4.1) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1290 | 20:26 |
| gitlab-br-bot | cs-shadow closed issue #1248 (Some public API functions can raise private LoadError exception) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1248 | 20:29 |
| *** cs-shadow has quit IRC | 22:41 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!