| *** bochecha has joined #buildstream | 01:00 | |
| *** bochecha has quit IRC | 01:02 | |
| *** bochecha has joined #buildstream | 01:02 | |
| *** bochecha has quit IRC | 01:06 | |
| *** bochecha has joined #buildstream | 01:07 | |
| *** bochecha has quit IRC | 01:12 | |
| *** bochecha has joined #buildstream | 01:22 | |
| *** bochecha has quit IRC | 01:27 | |
| *** bochecha has joined #buildstream | 01:27 | |
| *** bochecha has quit IRC | 01:31 | |
| *** bochecha has joined #buildstream | 01:52 | |
| *** bochecha has quit IRC | 01:56 | |
| *** bochecha has joined #buildstream | 02:11 | |
| *** bochecha has joined #buildstream | 02:12 | |
| *** bochecha has quit IRC | 02:16 | |
| *** bochecha has joined #buildstream | 02:17 | |
| *** bochecha has quit IRC | 02:27 | |
| *** bochecha has joined #buildstream | 02:32 | |
| *** bochecha has quit IRC | 02:53 | |
| *** bochecha has joined #buildstream | 03:41 | |
| *** bochecha has quit IRC | 03:42 | |
| *** palasso has joined #buildstream | 05:22 | |
| gitlab-br-bot | push on buildstream@deptype (by Jürg Billeter): 1 commit (last: _loader.py: Support dependency type 'all') https://gitlab.com/BuildStream/buildstream/commit/db49248dcc3480eb9f89cea84e6000cea763cbf2 | 05:57 |
|---|---|---|
| gitlab-br-bot | push on buildstream@deptype (by Jürg Billeter): 1 commit (last: tests/loader: Test dependency type 'all') https://gitlab.com/BuildStream/buildstream/commit/8260d0f5086bce840b1cb52c1675ac6038dc7224 | 06:08 |
| gitlab-br-bot | buildstream: issue #61 ("Add `all` value to depends `type`") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/61 | 06:15 |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 2 commits (last: _loader.py: Support dependency type 'all') https://gitlab.com/BuildStream/buildstream/commit/db49248dcc3480eb9f89cea84e6000cea763cbf2 | 06:15 |
| gitlab-br-bot | buildstream: Jürg Billeter deleted branch deptype | 06:16 |
| gitlab-br-bot | buildstream: issue #58 ("Catch exceptions when using OSTree version tracking in old OSTree versions") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/58 | 06:24 |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 1 commit (last: setup.py: Catch exception with OSTree < 2017.4) https://gitlab.com/BuildStream/buildstream/commit/67af5967041242a0392192aad285fbad6d0b44bb | 06:24 |
| *** bochecha has joined #buildstream | 07:04 | |
| *** bochecha has quit IRC | 07:09 | |
| *** bochecha has joined #buildstream | 07:09 | |
| *** bochecha_ has joined #buildstream | 07:17 | |
| *** bochecha_ has quit IRC | 07:18 | |
| *** bochecha_ has joined #buildstream | 07:19 | |
| *** bochecha has quit IRC | 07:19 | |
| *** bochecha_ is now known as bochecha | 07:19 | |
| *** anahuelamo has joined #buildstream | 07:41 | |
| gitlab-br-bot | push on buildstream@summary-status (by Jürg Billeter): 2 commits (last: _pipeline.py: Add remote_ticker for artifact list fetching) https://gitlab.com/BuildStream/buildstream/commit/df8b5a7d6a4a426cfd43d79bbf1cd10565442b2b | 07:46 |
| gitlab-br-bot | buildstream: merge request (summary-status->master: Print status message when fetching artifact list) #74 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/74 | 07:48 |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 1 commit (last: _pipeline.py: Fix --deps run) https://gitlab.com/BuildStream/buildstream/commit/1c2a884125092ef0bf9c53fc662de75099f77ec3 | 08:22 |
| gitlab-br-bot | buildstream: merge request (pip->master: Add a pip element) #75 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/75 | 09:07 |
| gitlab-br-bot | buildstream: merge request (pip->master: Add a pip element) #75 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/75 | 09:09 |
| *** ssam2 has joined #buildstream | 09:10 | |
| *** tlater has joined #buildstream | 09:20 | |
| *** anahuelamo has quit IRC | 09:30 | |
| *** palasso has quit IRC | 09:37 | |
| *** palasso has joined #buildstream | 09:39 | |
| *** bochecha_ has joined #buildstream | 09:54 | |
| *** bochecha has quit IRC | 09:56 | |
| *** bochecha_ is now known as bochecha | 09:57 | |
| *** bochecha has quit IRC | 10:02 | |
| *** bochecha has joined #buildstream | 10:04 | |
| *** anahuelamo has joined #buildstream | 10:11 | |
| *** anahuelamo has quit IRC | 10:31 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 3 commits (last: Clean _platform.py) https://gitlab.com/BuildStream/buildstream/commit/0b90c645baf3334338602957d850959510c873ab | 10:36 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 3 commits (last: Clean _platform.py) https://gitlab.com/BuildStream/buildstream/commit/2b666e5bb17938f72ee4ff38e74d883005bf6934 | 10:41 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 3 commits (last: Clean _platform.py) https://gitlab.com/BuildStream/buildstream/commit/a764c68cf18bae768e7716ac3dafdee48f2813b8 | 10:48 |
| *** anahuelamo has joined #buildstream | 10:49 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 3 commits (last: Clean _platform.py) https://gitlab.com/BuildStream/buildstream/commit/7b282a4fa7ac277cf6391e6012da5a76d0d9383e | 10:52 |
| *** anahuelamo has quit IRC | 11:48 | |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 1 commit (last: main.py: Fix crash with `bst workspace list` in bad directory) https://gitlab.com/BuildStream/buildstream/commit/ba71e3b96ea62124bfe248aabeb9c4922f0a4e3a | 11:58 |
| gitlab-br-bot | buildstream: issue #57 ("`bst workspace list` crashes in bad directory") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/57 | 11:58 |
| *** tlater has quit IRC | 12:18 | |
| *** anahuelamo has joined #buildstream | 12:21 | |
| *** anahuelamo has quit IRC | 12:24 | |
| *** anahuelamo has joined #buildstream | 12:29 | |
| *** tlater has joined #buildstream | 12:54 | |
| *** anahuelamo has quit IRC | 14:02 | |
| *** anahuelamo has joined #buildstream | 14:07 | |
| juergbi | tlater: source bundle is somewhat broken right now. see #59 but it looks like there are further issues | 14:14 |
| juergbi | i thought the idea was to generate a tarball that contains all build + their runtime dependencies + the source of the target package + a script to build the target package | 14:15 |
| juergbi | however, it seems to currently include sources+scripts for all build dependencies of the target package and their runtime packages. but not the target package itself, at least with default --deps build mode | 14:16 |
| juergbi | would you be able to shed some light on this? | 14:17 |
| tlater | Hm | 14:26 |
| tlater | Not sure, it should definitely include the target package, at least as far as I remember. | 14:27 |
| tlater | I'll have a look | 14:27 |
| juergbi | thanks | 14:27 |
| juergbi | added comment to #59 for the described issue. not pushing a fix before the whole feature makes sense to me | 14:31 |
| tlater | Yeah, that's alright. It's been a while, and I didn't understand exactly what build and run dependencies were back then, so it's entirely possible that the default is just wrong. | 14:32 |
| juergbi | default deps = 'build' actually makes sense to me, if we were to collect the artifacts instead of the sources of all dependencies | 14:34 |
| juergbi | my primary question is whether the goal is to build only the target element or whether the goal is to build a whole stack | 14:34 |
| tlater | The goal was to build a whole stack - this is designed to bootstrap from a stage2 | 14:35 |
| ssam2 | yeah the idea is the target element would be something like build-system.bst | 14:35 |
| juergbi | ok, in that case we actually need more than Scope.BUILD dependencies | 14:36 |
| tlater | It asks for Scope.ALL in the actual pipeline | 14:36 |
| juergbi | we also need the build dependencies of the build dependencies | 14:36 |
| tlater | I think the default for --deps is just wrong | 14:36 |
| tlater | Or rather, --deps makes no sense for this command | 14:36 |
| juergbi | right, i guess all would work fine | 14:37 |
| tlater | Hmm | 14:37 |
| tlater | There was something about script dependencies | 14:37 |
| *** ssam2 has quit IRC | 14:39 | |
| juergbi | right, need a way to execute the first script before the bootstrap element is assembled | 14:39 |
| tlater | No, Scope.ALL should work. Script dependencies were excluded completely from a source bundle | 14:39 |
| tlater | I don't quite recall the reasoning behind that, but that's how it's implemented. The deps flag should just be removed, otherwise it should be fine. | 14:40 |
| juergbi | i'm not sure whether the master build script is all fine yet but dependency wise Scope.ALL should be fine, yes | 14:42 |
| *** anahuelamo has quit IRC | 14:42 | |
| juergbi | just also need to cover ImportElement.generate_script to actually fix #59 | 14:42 |
| tlater | What's the problem with the master build script? | 14:42 |
| juergbi | i'm not aware of a specific issue, just haven't looked at it more closely. don't know whether it properly handles staging paths for dependencies and build outputs etc. | 14:43 |
| tlater | I think it does, but it could use some testing. Should I check that while I wait for AIX builds or are you already on it? | 14:44 |
| juergbi | i can take a look now that i know how it's supposed to work | 14:45 |
| tlater | Alright, thanks :) | 14:45 |
| *** ssam2 has joined #buildstream | 14:55 | |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 1 commit (last: main.py: Include all dependencies in source bundle) https://gitlab.com/BuildStream/buildstream/commit/084b50420c27c28291ea110466bf0de8b18073f7 | 14:57 |
| *** anahuelamo has joined #buildstream | 15:05 | |
| juergbi | with #59 fixed (ImportElement) i expect it to basically work, however, as it is, there will definitely be corner cases where build may fail or differ from buildstream-driven build | 15:09 |
| *** anahuelamo has quit IRC | 15:09 | |
| *** anahuelamo has joined #buildstream | 15:09 | |
| juergbi | it copies the install-root of every built element into / | 15:09 |
| juergbi | i.e., it doesn't necessarily stage them the same way as bst | 15:10 |
| juergbi | could be an issue with conflicting elements or with configure scripts that autodetect libraries (so it behaves differently when additional libraries are staged) | 15:11 |
| juergbi | and the use of some external 'stage2' as bootstrap can also influence the build, though. and that simply gets overwritten more and more | 15:12 |
| tlater | juergbi: Tristan decided we didn't want to focus too much on that, and that that was good enough, but yeah, this was a discussion. Perhaps leave an issue? | 15:12 |
| tlater | The stage2 is intentionally overwritten though - The stage2 binaries are supposed to be replaced by natively built elements. | 15:13 |
| juergbi | it would be nice to get closer to regular bst behavior but there might well be more pressing matters | 15:13 |
| juergbi | yes, just mentioning it as potential difference | 15:13 |
| juergbi | i'm wondering whether we couldn't simply place the source of the lowest level element of the stack (pretty much has to be an import element) in the root of the tarball | 15:14 |
| juergbi | ideally, we would probably even use chroot in the script for each element | 15:14 |
| juergbi | and stage dependencies the same way as regular bst | 15:14 |
| juergbi | integration commands also appear to be missing | 15:15 |
| tlater | The scripts call ldconfig after each build instead | 15:15 |
| juergbi | yes, i saw that. may or may not be sufficient | 15:16 |
| juergbi | if it's just meant for bootstrapping a new system, it might be ok | 15:16 |
| tlater | Yup, definitely. | 15:16 |
| tlater | As in, might not be sufficient | 15:17 |
| tlater | x) | 15:17 |
| tlater | juergbi: Can the build cache key and the WEAK cache key ever overlap? | 15:26 |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 1 commit (last: tarcache.py: Add tarfile fallbacks) https://gitlab.com/BuildStream/buildstream/commit/ba35d0d940713952329365f1a1ad52749eff4b43 | 16:02 |
| *** ssam2 has quit IRC | 17:00 | |
| gitlab-br-bot | push on buildstream@cross_platform (by Tristan Maat): 20 commits (last: Add `bst pull` command) https://gitlab.com/BuildStream/buildstream/commit/39696b17ec1410f0110e5c3c7a0510a8151b57ac | 17:06 |
| *** anahuelamo has quit IRC | 17:38 | |
| bochecha | hopefully I've got the pip element tests working now :) | 17:55 |
| bochecha | running the complete set of tests, just to make sure I didn't break anything | 17:55 |
| juergbi | tlater: yes, mainly when there are no build dependencies, the two cache keys are the same | 18:04 |
| juergbi | bochecha: sorry about the misleading comment, i forgot that it was auto-generated | 18:05 |
| gitlab-br-bot | buildstream: merge request (pip->master: Add a pip element) #75 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/75 | 18:05 |
| bochecha | juergbi: no problem :) | 18:05 |
| bochecha | just figured it out while the tests are running | 18:05 |
| bochecha | juergbi: I just pushed to the branch, it should have all your comments fixed | 18:06 |
| bochecha | juergbi: and thanks for having reviewed so quickly :) | 18:06 |
| *** tlater has quit IRC | 18:08 | |
| *** bochecha_ has joined #buildstream | 18:10 | |
| *** bochecha has quit IRC | 18:12 | |
| *** bochecha_ is now known as bochecha | 18:12 | |
| gitlab-br-bot | push on buildstream@master (by Jürg Billeter): 2 commits (last: makemaker.py: Fix typo in comment) https://gitlab.com/BuildStream/buildstream/commit/5e70714f963013c3ccab8e084ca7adda781f3618 | 18:50 |
| gitlab-br-bot | buildstream: merge request (pip->master: Add a pip element) #75 changed state ("closed"): https://gitlab.com/BuildStream/buildstream/merge_requests/75 | 18:51 |
| *** bochecha_ has joined #buildstream | 19:01 | |
| *** bochecha has quit IRC | 19:01 | |
| *** bochecha_ is now known as bochecha | 19:02 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!