| *** alatiera has quit IRC | 01:53 | |
| *** tristan has quit IRC | 06:34 | |
| *** tristan has joined #buildstream | 06:40 | |
| *** tristan has quit IRC | 08:29 | |
| *** tpollard has joined #buildstream | 08:46 | |
| *** raoul has joined #buildstream | 08:54 | |
| benschubert | raoul: are you happy with https://gitlab.com/BuildStream/buildstream/merge_requests/1350#note_174283803 ? | 09:03 |
|---|---|---|
| raoul | Yep that looks good :) | 09:04 |
| *** tristan has joined #buildstream | 09:05 | |
| benschubert | great thanks! Anything else from your side on the MR? Otherwise I'll soon merge it | 09:05 |
| benschubert | tristan: concerning https://gitlab.com/BuildStream/website/merge_requests/121 should we wait for the split to happen before merging? Or should we add it now? | 09:05 |
| *** ChanServ sets mode: +o tristan | 09:06 | |
| tristan | benschubert, I'm fine merging it without the added cython dependency mention, as the rest of it is not exactly untrue for 1.2 | 09:06 |
| benschubert | ok! I'll update it like that then | 09:07 |
| tristan | benschubert, however for what is the main user facing instructions on the website, I'd rather keep it all 1.2 centric until we can split the 1.2/master docs indeed | 09:07 |
| tristan | sure :) | 09:07 |
| benschubert | yep so not having cython makes sense :) | 09:07 |
| benschubert | great, that means I'll be able to merge the cython today o/ | 09:08 |
| benschubert | If someone wants to have a look at it, please let me know before I hit merge | 09:08 |
| tristan | Yeah, the website is not exactly blocking that anyway :) | 09:08 |
| *** phil has joined #buildstream | 09:09 | |
| * tristan steps out to meet folks for dinner | 09:09 | |
| raoul | LGTM benschubert, the remaining discussions look resolved but I'll leave that up to the people who raised them to decide if they're happy | 09:09 |
| benschubert | Thanks! | 09:09 |
| *** jonathanmaw has joined #buildstream | 09:22 | |
| gitlab-br-bot | marge-bot123 merged MR !1350 (bschubert/cython->master: Introduce Cython for better performances) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1350 | 09:34 |
| benschubert | ^ we have cython now, welcome the build step when developing :) | 09:35 |
| gitlab-br-bot | aevri opened (was WIP) MR !1334 (aevri/split_jobs_parent_child->master: Split ChildJob out from Job class) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1334 | 09:36 |
| Kinnison | woot! | 09:36 |
| * Kinnison rebases his branch and retests his in-progress chance | 09:38 | |
| Kinnison | change | 09:38 |
| *** lachlan has joined #buildstream | 09:40 | |
| *** lachlan has quit IRC | 09:47 | |
| tpollard | do I need to 'pip3 install --user .' every time I make a local change now? | 10:07 |
| gitlab-br-bot | danielsilverstone-ct opened MR !1364 (danielsilverstone-ct/iterative-circdeps->master: Rewrite `Loader._check_circular_deps()` to be iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1364 | 10:09 |
| Kinnison | pip3 install -e . no? | 10:10 |
| Kinnison | (and you only *have* to do it if you change a .pyx file IIRC) | 10:10 |
| tpollard | I was looking at https://gitlab.com/BuildStream/website/merge_requests/121/diffs | 10:12 |
| tpollard | which says it won't work? | 10:13 |
| *** lachlan has joined #buildstream | 10:14 | |
| Kinnison | benschubert: ^^^ ? | 10:20 |
| benschubert | tpollard: if you are on master, yes | 10:22 |
| benschubert | only for cython files | 10:22 |
| benschubert | or wehenever you rebase on master/change branch, if and only if a cython file has changed | 10:22 |
| benschubert | does that make sense? | 10:23 |
| * tpollard nods | 10:23 | |
| tpollard | thanks benschubert :) | 10:23 |
| benschubert | tpollard: however if oyu use tox when testing, you won't need to do that, tox will do it for you :) | 10:23 |
| *** lachlan has quit IRC | 10:24 | |
| Kinnison | benschubert: is that true even if you don't pass -r ? | 10:24 |
| benschubert | Kinnison: yes! | 10:25 |
| Kinnison | Nice! | 10:25 |
| benschubert | tox has been changed to work with a wheel of the project and the wheel is created on invocation | 10:25 |
| benschubert | which arguably slows down the process a bit | 10:25 |
| benschubert | but ensure everything is correct | 10:25 |
| Kinnison | nod. | 10:27 |
| *** toscalix has joined #buildstream | 10:30 | |
| gitlab-br-bot | BenjaminSchubert approved MR !1364 (danielsilverstone-ct/iterative-circdeps->master: Rewrite `Loader._check_circular_deps()` to be iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1364 | 10:31 |
| *** lachlan has joined #buildstream | 10:33 | |
| *** cs-shadow has quit IRC | 10:46 | |
| *** cs-shadow has joined #buildstream | 10:47 | |
| *** lachlan has quit IRC | 10:48 | |
| Kinnison | raoul: any idea why this failure? https://gitlab.com/BuildStream/buildstream/-/jobs/222208879 (doesn't happen on other platforms) | 10:55 |
| raoul | was just looking at that, happens on one of my branches | 10:55 |
| raoul | https://gitlab.com/BuildStream/buildstream/-/jobs/221730468 | 10:56 |
| raoul | no idea so far | 10:56 |
| Kinnison | The message appears to be "failed to connect to all addresses" | 10:56 |
| *** lachlan has joined #buildstream | 10:56 | |
| Kinnison | Rather than "Connect Failed" | 10:56 |
| Kinnison | I wonder if it's a libc error message change | 10:56 |
| raoul | Yeah which doesn't seem to appear in the buildstream codebase which is confusing | 10:57 |
| raoul | that's plausible | 10:57 |
| raoul | Is that image on newer libraries than the rest? | 10:57 |
| Kinnison | dunno | 10:57 |
| tpollard | ' # Check if the tests pass after updating requirements to their latest allowed version.' | 11:01 |
| raoul | looks like it might be a grpc thing | 11:01 |
| *** lachlan has quit IRC | 11:01 | |
| tpollard | sadness | 11:06 |
| raoul | that is what I associate with grpc yes | 11:07 |
| raoul | yeah looks like it's definitely a grpc thing, could expand the test to allow that error message? | 11:10 |
| tpollard | https://github.com/grpc/grpc/blob/master/src/core/ext/filters/client_channel/lb_policy/pick_first/pick_first.cc | 11:12 |
| raoul | yeah I found that too, that error message isn't present in versions of grpc that we're using in other images | 11:13 |
| tpollard | keep associating to sadness | 11:13 |
| gitlab-br-bot | BenjaminSchubert opened issue #1039 (Tracker for hotspots to move to cython) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1039 | 11:21 |
| *** lachlan has joined #buildstream | 11:24 | |
| gitlab-br-bot | marge-bot123 merged MR !1364 (danielsilverstone-ct/iterative-circdeps->master: Rewrite `Loader._check_circular_deps()` to be iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1364 | 11:25 |
| *** slaf has quit IRC | 11:31 | |
| *** slaf has joined #buildstream | 11:32 | |
| *** slaf has joined #buildstream | 11:32 | |
| *** slaf has joined #buildstream | 11:33 | |
| *** slaf has joined #buildstream | 11:33 | |
| *** slaf has joined #buildstream | 11:33 | |
| *** slaf has joined #buildstream | 11:34 | |
| *** slaf has joined #buildstream | 11:34 | |
| *** lachlan has quit IRC | 11:53 | |
| gitlab-br-bot | danielsilverstone-ct opened MR !1365 (danielsilverstone-ct/iterative-loader-bits->master: Make more parts of the loader iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1365 | 12:44 |
| *** finn has joined #buildstream | 12:59 | |
| Kinnison | jennis: you have two pipelines on the same branch -- do youwant the older one killing? | 13:11 |
| tpollard | I want a groundskeeper willie bot to garden the pipelines | 13:12 |
| Kinnison | heh | 13:13 |
| jennis | Kinnison, yeah kill the spare | 13:16 |
| Kinnison | Yes jennisdamort | 13:17 |
| jennis | I'm glad you got that | 13:17 |
| gitlab-br-bot | BenjaminSchubert approved MR !1365 (danielsilverstone-ct/iterative-loader-bits->master: Make more parts of the loader iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1365 | 13:38 |
| *** lachlan has joined #buildstream | 13:38 | |
| gitlab-br-bot | willsalmon closed MR !1353 (shashwatdalal/updatesiteflags->master: tests: Change HAVE_SANDBOX to fail on windows/OSX) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1353 | 13:42 |
| *** kapip has joined #buildstream | 13:43 | |
| gitlab-br-bot | BenjaminSchubert reopened MR !1353 (shashwatdalal/updatesiteflags->master: tests: Change HAVE_SANDBOX to fail on windows/OSX) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1353 | 13:47 |
| *** lachlan has quit IRC | 13:49 | |
| *** lachlan has joined #buildstream | 14:11 | |
| *** lachlan has quit IRC | 14:21 | |
| *** lachlan has joined #buildstream | 14:29 | |
| *** lachlan has quit IRC | 14:32 | |
| gitlab-br-bot | jjardon reopened issue #560 (BUG: Message handling out of sync) on buildstream https://gitlab.com/BuildStream/buildstream/issues/560 | 14:33 |
| gitlab-br-bot | jjardon closed issue #560 (BUG: Message handling out of sync) on buildstream https://gitlab.com/BuildStream/buildstream/issues/560 | 14:34 |
| gitlab-br-bot | jjardon opened issue #1040 (BUG: Message handling out of sync) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1040 | 14:35 |
| *** lachlan has joined #buildstream | 14:43 | |
| *** lachlan has quit IRC | 14:47 | |
| benschubert | phil: We have https://gitlab.com/BuildStream/buildstream/blob/master/tests/testutils/site.py and https://gitlab.com/BuildStream/buildstream/blob/master/src/buildstream/testing/_utils/site.py which seem quite similar in intent. Is there a reason we hadn't merged them together when extracting the testing stuff for plugins? | 14:57 |
| phil | benschubert, I don't see any reason why they shouldn't have been merged so I think it was an oversight | 15:02 |
| *** lachlan has joined #buildstream | 15:02 | |
| benschubert | phil: Ok, so we should go for a single testing/_utils/site.py right? | 15:02 |
| phil | That would make sense to me | 15:03 |
| benschubert | thanks! | 15:03 |
| phil | :) | 15:03 |
| *** lachlan has quit IRC | 15:08 | |
| gitlab-br-bot | marge-bot123 merged MR !1365 (danielsilverstone-ct/iterative-loader-bits->master: Make more parts of the loader iterative) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1365 | 15:24 |
| *** lachlan has joined #buildstream | 15:26 | |
| *** lachlan has quit IRC | 15:39 | |
| *** lachlan has joined #buildstream | 15:42 | |
| benschubert | phil: also the _utils.site is private in buildstream/testing. Is that also an oversite or do we actually don't want other projects to get access to it? | 16:02 |
| phil | benschubert, I made it private at the time to err on the side of conservatism with what I was exposing. If there's desire to have it exposed, then I have no objections. | 16:07 |
| benschubert | I think having access to HAS_SANDBOX and various others is sueful for integration testing of projects | 16:08 |
| benschubert | but I'm not sure if we should open everythin | 16:09 |
| benschubert | what about having testing/site.py in which we add things that are publicly available? | 16:10 |
| phil | seems reasonable to me. I agree that it doesn't make sense to make it all public. HAVE_SANDBOX would make sense to be public, and potentially IS_{LINUX|WINDOWS|WSL}. | 16:16 |
| benschubert | agreed, I'll move those there | 16:18 |
| benschubert | thanks! | 16:18 |
| *** alatiera has joined #buildstream | 16:29 | |
| *** raoul has quit IRC | 16:34 | |
| *** lachlan has quit IRC | 16:35 | |
| *** lachlan has joined #buildstream | 16:37 | |
| finn | Is there any way to configure bubblewrap when running a buildstream build? | 16:41 |
| benschubert | finn: not sure what you mean by that, what do you need to do? | 16:41 |
| finn | Play around with some of the flags bwrap uses when executing an action | 16:42 |
| finn | e.g when it's running configure for an autotools project | 16:43 |
| benschubert | finn: unfortunately there is no way from a project or the cli to configure that. You could hack the parameters in sandbox/_sandboxbwrap.py if you want to experiment | 16:44 |
| finn | Thought so, just wanted to double check :) | 16:44 |
| finn | I'm playing around with Buildstream v1.2 and I currently get: | 16:45 |
| finn | `bwrap: execvp sh: Permission denied` | 16:45 |
| finn | When running the autotools Buildstream example during the configure stage. My environment is a little strange; I'm running things inside a `linux-sandbox` inside a docker container. When I run Buildstream in my docker container, outside the `linux-sandbox`, I don't get this error. | 16:45 |
| benschubert | would running your docker with "--privileged" work? | 16:48 |
| benschubert | If you, your docker is missing a permission | 16:48 |
| finn | tried that | 16:48 |
| finn | hence why it works when I run buildstream in the container, just not in the `linux-sandbox` | 16:48 |
| finn | Interestingly, if I run a simple bwrap command inside that environment, it works. Though I don't have as many flags set as Buildstream does. | 16:49 |
| finn | Thanks anyway :) | 16:54 |
| benschubert | I've never used linux-sandbox, won't be of much help, sorry :/ | 16:59 |
| finn | Neither have I ^ - ^ | 16:59 |
| *** finn has quit IRC | 17:06 | |
| *** jonathanmaw has quit IRC | 17:31 | |
| *** finn has joined #buildstream | 17:45 | |
| *** lachlan has quit IRC | 17:47 | |
| *** finn has quit IRC | 18:09 | |
| *** finn has joined #buildstream | 18:45 | |
| *** xjuan has joined #buildstream | 19:23 | |
| *** finn has quit IRC | 19:42 | |
| *** xjuan has quit IRC | 20:20 | |
| *** toscalix has quit IRC | 20:33 | |
| *** bakhtyar has joined #buildstream | 20:42 | |
| *** bakhtyar has quit IRC | 20:43 | |
| *** xjuan has joined #buildstream | 20:49 | |
| *** xjuan has quit IRC | 21:24 | |
| *** xjuan has joined #buildstream | 21:41 | |
| *** xjuan has quit IRC | 22:01 | |
| *** alatiera has quit IRC | 22:50 | |
| *** dftxbs3e has quit IRC | 23:44 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!