| *** mcatanzaro has joined #buildstream | 00:01 | |
| *** tristan has quit IRC | 00:08 | |
| *** Prince781 has quit IRC | 00:11 | |
| *** Prince781 has joined #buildstream | 01:12 | |
| *** mcatanzaro has quit IRC | 05:24 | |
| *** Prince781 has quit IRC | 06:53 | |
| *** valentind has joined #buildstream | 08:30 | |
| *** dominic has joined #buildstream | 08:33 | |
| *** Prince781 has joined #buildstream | 08:45 | |
| *** tristan has joined #buildstream | 08:50 | |
| *** aday has joined #buildstream | 08:59 | |
| *** valentind has quit IRC | 09:05 | |
| *** toscalix has joined #buildstream | 09:12 | |
| *** ernestask has joined #buildstream | 09:19 | |
| *** Prince781 has quit IRC | 09:23 | |
| *** ssam2 has joined #buildstream | 10:00 | |
| *** jonathanmaw has joined #buildstream | 10:05 | |
| ssam2 | I had a go at sketching out a BuildStream quick start guide: https://wiki.gnome.org/SamThursfield/BuildStreamQuickStart | 10:43 |
|---|---|---|
| ssam2 | feedback welcome, although i don't guarantee spending more time on it myself | 10:44 |
| ssam2 | but i had some ideas about how to structure it and such that i wanted to get out of my head :-) | 10:45 |
| jmac | Looks good, I'll have a proper read through later | 10:46 |
| aiden | why `bst` and not `bs`? | 10:47 |
| ssam2 | LOL | 10:47 |
| aiden | `bs` might stick a bit more | 10:47 |
| ssam2 | 'bs' has some ... negative connotations | 10:48 |
| persia | aiden: Unfortunate connotations to "bs", also memetic heritage encouraging three letter filename extensions. | 10:48 |
| aiden | `git` also has negative connotations | 10:48 |
| skullman | less negative than `bs` | 10:48 |
| persia | Yes, but the author of git is famous for not caring if other folk are offended. That luxury does not apply to most folk. | 10:49 |
| jmac | "bst" conjures up images of long summer evenings if you're in Britain | 10:50 |
| nexus | that was the name i gave to my remote for a while :p | 10:50 |
| *** theawless[m] has joined #buildstream | 10:50 | |
| tristan | I came up with the BST_ namespaced inspired from gstreamer and GST_ fwiw; it seems that only my close colleagues on this project had ever thought up `bs`, aiden gets an award :) | 10:50 |
| aiden | it would make me smile if i had to type `bs` more than `bst` | 10:51 |
| aiden | just saying | 10:52 |
| tristan | heh | 10:52 |
| gitlab-br-bot | buildstream: merge request (image-authoring->master: WIP: Image authoring documentation) #262 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/262 | 11:30 |
| gitlab-br-bot | buildstream: merge request (image-authoring->master: WIP: Image authoring documentation) #262 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/262 | 11:30 |
| gitlab-br-bot | buildstream: merge request (image-authoring->master: WIP: Image authoring documentation) #262 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/262 | 11:30 |
| *** jennis has quit IRC | 11:32 | |
| *** jennis has joined #buildstream | 11:34 | |
| gitlab-br-bot | buildstream: merge request (buildproject->master: Added buildproject doc) #238 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/238 | 11:36 |
| *** toscalix has quit IRC | 11:44 | |
| *** toscalix has joined #buildstream | 11:45 | |
| gitlab-br-bot | buildstream: merge request (image-authoring->master: WIP: Image authoring documentation) #262 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/262 | 11:45 |
| *** toscalix has quit IRC | 11:47 | |
| *** toscalix has joined #buildstream | 11:52 | |
| jmac | Hmm, where do I go with this error? https://paste.gnome.org/p9ibo5nln | 12:00 |
| ssam2 | that's some error in variable resolution which isn't reported very clearly | 12:01 |
| ssam2 | i imagine you used a variable that isn't defined in an element | 12:01 |
| ssam2 | and buildstream reports the last place it looked for it, instead of where it actually came from | 12:01 |
| ssam2 | i've had this before, but always been in too much of an excited rush to actually file an issue | 12:02 |
| tristan | oooh nice catch; that would be good to fix | 12:05 |
| nexus | juergbi: ok, i've added the fix and written a test for it, going to push the MR shortly | 12:20 |
| juergbi | great | 12:24 |
| gitlab-br-bot | buildstream: merge request (sam/plugin-error->master: _plugincontext.py: Fix unhandled exception if plugin is not found) #274 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/274 | 12:28 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: WIP: Issue 166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 12:31 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: Issue 166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 12:34 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: Issue #166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 13:16 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: Issue #166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 13:16 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: Issue #166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 13:17 |
| nexus | juergbi: i've updated my commit message and added the file. However my CI is still failing and is up to date with master | 13:29 |
| juergbi | nexus: it still says '1 commit behind' here | 13:29 |
| juergbi | and the summary line of the commit message is still: This fixes #166 and adds tests | 13:30 |
| gitlab-br-bot | buildstream: merge request (issue-166_yaml_removing_underscores->master: Issue #166 yaml removing underscores) #245 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/245 | 13:35 |
| persia | nexus: I thought the new test in context.py was going to test fetchers: "twenty-two" | 14:04 |
| persia | I also thought the 1.23.4 and 1_23_4 were going to be new tests in yaml.py | 14:05 |
| *** tristan has quit IRC | 14:06 | |
| juergbi | i already commented on the MR | 14:07 |
| persia | And your comments are much more in-depth and useful than mine :) | 14:10 |
| juergbi | nexus: btw: i suggest using branches in the main repo in the future as that will use our CI runner which seems to be faster (and reliably uses a cache) | 14:10 |
| *** tristan has joined #buildstream | 14:18 | |
| *** toscalix has quit IRC | 14:19 | |
| *** toscalix has joined #buildstream | 14:21 | |
| *** mcatanzaro has joined #buildstream | 14:29 | |
| nexus | When implementing my tests, there is currently one that fails "string parsing for floats". Should i have this test skipped or have it as an expected failure, or have it expect 1.2 and not 1.20 (basically testing that it still doesn't work) <- not a fan of the last one | 14:36 |
| nexus | juergbi: | 14:37 |
| nexus | juergbi: i can't create branches in the main repo | 14:37 |
| persia | nexus: Hrm? What is the failure on parsing 1.20 as a string and getting 1.20 as a string? | 14:38 |
| juergbi | nexus: i would suggest also fixing the float case at the same time. if it's more work than just adding add_constructor() for floats and you want to handle this as a separate merge request, i would move the float test also to that other MR | 14:39 |
| nexus | kk | 14:41 |
| juergbi | nexus: i just added your user to the buildstream group. you can push new branches for future MRs | 14:41 |
| nexus | woo \o/ | 14:42 |
| gitlab-br-bot | buildstream: issue #253 ("Unhelpful error message on missing variables") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/253 | 14:55 |
| nexus | hm. that was easy | 14:57 |
| nexus | juergbi: re: the functionality causes quotes to be added to the track value. | 15:05 |
| nexus | As we discussed earlier today, would it not be nicer to keep this, and in the docs, encourage them to be used to prevent this issue, whilst allowing people to not do so without it breaking stability? | 15:05 |
| nexus | This way, we can phase in the use of quotes for tracking values, without forcing the user to do so on existing projects | 15:05 |
| juergbi | nexus: it does also convert double quotes to single quotes, though, afaict | 15:06 |
| nexus | does that matter? | 15:06 |
| juergbi | mainly just about diff noise | 15:07 |
| juergbi | not sure how much of an inconvenience it would be in practice | 15:07 |
| juergbi | we might need to check with tristan what he thinks about this | 15:07 |
| nexus | wouldn't it only happen once? | 15:07 |
| nexus | kk | 15:08 |
| juergbi | if the change is committed, yes | 15:08 |
| juergbi | which means that it might be ok in practice but i still think it's not ideal | 15:08 |
| juergbi | i assume it will do this also for unrelated values, not just for track | 15:08 |
| juergbi | so it might be more than just one line | 15:09 |
| juergbi | yes, it will quote all values that could be interpreted as non-strings in the file | 15:11 |
| juergbi | (and change quotes to single where currently double quotes are used) | 15:12 |
| juergbi | i would definitely check whether we can easily avoid this | 15:12 |
| juergbi | if it turns out to be a lot of work, we can re-evaluate | 15:12 |
| ssam2 | we might as well just switch to https://github.com/imbal/safeyaml at that point | 15:13 |
| nexus | Wouldn't it be bad to have to convert them every time? Wouldn't it be faster if they were all just read as strings initially?> | 15:13 |
| juergbi | ssam2: hm? | 15:14 |
| ssam2 | for a computer to convert an integer to a string? that takes nanoseconds ... | 15:14 |
| juergbi | nexus: with this MR, we do read them as strings | 15:14 |
| juergbi | it never even goes via integer | 15:14 |
| nexus | ssam2: kk | 15:14 |
| ssam2 | i mean, if we convert all non-string values to strings, we are starting to reimplement the "SafeYAML" subset of YAML documented at https://github.com/imbal/safeyaml | 15:14 |
| juergbi | ssam2: ah ok | 15:14 |
| juergbi | so i take it you're also in favor of fixing this issue | 15:15 |
| ssam2 | which may be a good idea, but not something we should sneak in by the back door | 15:15 |
| juergbi | if this were pre-1.0 days, i'd say it makes sense to discuss this. now, i'd say it's too late unless we hit serious roadblocks with the current approach | 15:15 |
| ssam2 | yes... i mean, `bst track` is still a neat thing, but also having computers and people editing the same file is a recipe for madness | 15:15 |
| ssam2 | but if we are going to try to edit files the user wrote, we should do it properly, not decide to rewrite stuff for them in different ways | 15:16 |
| nexus | simple, lets just replace people with machines | 15:16 |
| *** cs_shadow has joined #buildstream | 15:22 | |
| tristan | ssam2, agreed; diffs are "bad", and you might be surprised to know how many times I've received bugs along the lines of "Glade completely changed the .ui file I created by hand, bad bad diffs !!" | 15:34 |
| tristan | while I think in that context the expectation is unreasonable, it could totally be done and would be nice | 15:34 |
| *** Prince781 has joined #buildstream | 15:49 | |
| *** dominic has quit IRC | 15:53 | |
| *** dominic has joined #buildstream | 15:53 | |
| persia | ssam2: Do we expect any yaml bahviours that are not implemented by SafeYAML? Are SafeYAML dependencies small enough to be a sensible alternative to ruamel for us? | 15:56 |
| ssam2 | SafeYAML is a subset of the YAML standard | 15:57 |
| ssam2 | so yes, we expect behaviours that it doesn't implement | 15:57 |
| ssam2 | or rather, explicitly forbids | 15:57 |
| skullman | having had a quick look at the ruamel code, it should be possible to preserve quoting style so it doesn't get normalised to single quotes by passing preserve_quotes=True to ruamel.yaml.load | 15:59 |
| juergbi | the issue about failing to preserve the quoting style is already present in master. e.g., it removes quotes if they are not necessary | 16:05 |
| juergbi | so it wouldn't be a regression of !245 and i'd be fine handling this separately (if at all) | 16:06 |
| juergbi | nexus: ^^ | 16:06 |
| * juergbi should rather add this to the MR | 16:06 | |
| gitlab-br-bot | buildstream: merge request (212-git-source-needs-a-way-to-disable-checking-out-submodules->master: Resolve "Git source needs a way to disable checking out submodules") #259 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/259 | 16:28 |
| *** toscalix has quit IRC | 16:44 | |
| nexus | 15:59 <skullman > having had a quick look at the ruamel code, it should be possible to preserve quoting style so it doesn't get normalised to single quotes by passing preserve_quotes=True to ruamel.yaml.load | 17:07 |
| nexus | juergbi: this worked | 17:07 |
| juergbi | nice | 17:07 |
| nexus | you happy for this to be pushed? | 17:08 |
| juergbi | nexus: you haven't updated the branch yet, have you? | 17:14 |
| juergbi | i'm certainly happy to add preserve_quotes in the same MR. although i would put it in a separate commit | 17:14 |
| nexus | no | 17:14 |
| nexus | ok | 17:15 |
| nexus | is there an issue for this? | 17:17 |
| ssam2 | this has been discussed to death so much that i know it from memory ... | 17:20 |
| ssam2 | #166 | 17:20 |
| nexus | i mean for preserving quotes :) | 17:30 |
| juergbi | no, i don't think so | 17:30 |
| *** valentind has joined #buildstream | 17:34 | |
| *** Prince781 has quit IRC | 17:40 | |
| *** tiago has quit IRC | 17:48 | |
| *** jonathanmaw has quit IRC | 17:56 | |
| *** jonathanmaw has joined #buildstream | 18:02 | |
| *** Prince781 has joined #buildstream | 18:07 | |
| *** jonathanmaw has quit IRC | 18:18 | |
| *** dominic has quit IRC | 18:33 | |
| *** ssam2 has quit IRC | 18:43 | |
| *** Prince781 has quit IRC | 18:49 | |
| *** ernestask has quit IRC | 20:49 | |
| *** mcatanzaro has quit IRC | 21:16 | |
| *** Prince781 has joined #buildstream | 22:29 | |
| *** valentind has quit IRC | 22:51 | |
| *** tristan has quit IRC | 23:29 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!