| *** raoul has quit IRC | 00:59 | |
| *** raoul has joined #buildstream | 00:59 | |
| *** catonano has joined #buildstream | 02:59 | |
| *** alatiera__ has quit IRC | 04:09 | |
| *** catonano has quit IRC | 05:21 | |
| *** catonano has joined #buildstream | 06:06 | |
| juergbi | Kinnison: I see the issue but looking at the whole thing a bit more, I want to change a lot more about how platform is handled | 06:07 |
|---|---|---|
| juergbi | we can possibly completely rip out context from platform | 06:08 |
| juergbi | which would make lazy initialization trivial | 06:08 |
| juergbi | but it requires a bit of work | 06:08 |
| Kinnison | That would be nicer I agree | 06:46 |
| * Kinnison is a little sad the "quick fix" didn't ameliorate things though :/ | 06:51 | |
| juergbi | without the assert my local branch might actually work | 06:53 |
| juergbi | however, it seems a bit hacky | 06:53 |
| Kinnison | Yeah I'm not sure I'd want to drop the assert | 06:53 |
| Kinnison | it'd be *wrong* | 06:53 |
| juergbi | I could try the nicer approach in the next couple of hours | 06:54 |
| Kinnison | If you get a chance, that'd be cool | 06:54 |
| *** mohan43u has quit IRC | 07:00 | |
| *** raoul has quit IRC | 07:01 | |
| *** raoul has joined #buildstream | 07:03 | |
| *** iker has joined #buildstream | 07:07 | |
| *** mohan43u has joined #buildstream | 07:13 | |
| *** raoul has quit IRC | 07:15 | |
| *** abderrahim has quit IRC | 07:17 | |
| *** catonano has quit IRC | 07:41 | |
| *** catonano has joined #buildstream | 07:41 | |
| adds68 | valentind, i'll push a branch and link it to you :) | 07:53 |
| *** toscalix has joined #buildstream | 07:56 | |
| *** bochecha has joined #buildstream | 08:26 | |
| *** catonano has quit IRC | 09:00 | |
| tiagogomes | Can we fix whatever is wrong in the buildstream GNOME wiki so that it doesn't show "Build Stream" here: https://wiki.gnome.org/RecentChanges | 09:06 |
| toscalix | tiagogomes: funny, I do not see such separation here: https://wiki.gnome.org/Projects/BuildStream | 09:15 |
| toscalix | I guess it has to do with two capital letters together | 09:16 |
| toscalix | noidea about why | 09:16 |
| jmac | I would guess it's because Gnome wiki still uses camel case for page names and it's an attempt to unencode "BuildStream" | 09:17 |
| jmac | I doubt it's fixable without changing the whole wiki | 09:18 |
| tlater[m] | jjardon: Sorry, I missed the messages yesterday - my account name is tlater, unsurprisingly | 09:28 |
| gitlab-br-bot | buildstream: merge request (mac_fixes->master: Implement compatibility fixes for MacOSX and WSL Blocks #411 and #412") #726 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/726 | 09:29 |
| juergbi | Kinnison: I should probably move the new commits to a separate MR but first have to see whether that fully works | 09:35 |
| juergbi | Nexus: in case you're interested, I've pushed a preliminary update to mac_fixes with a whole set of new commits underneath | 09:36 |
| * tlater[m] wonders how hard it would be to implement a BuildStream remote execution client in a web browser | 09:54 | |
| tlater[m] | You could have a little badge under your GitHub project that builds it... | 09:55 |
| tlater[m] | juergbi: Do you remember when junctions where added proper? Was this before or after 1.0.1? | 10:09 |
| tlater[m] | s/where/were/ | 10:09 |
| * tlater[m] is trying to infer from MRs and git commits, but there were a bunch of them... | 10:12 | |
| *** ahmed89 has joined #buildstream | 10:12 | |
| tlater[m] | Ah, 1.1.0~27! | 10:13 |
| valentind | juergbi, It seems that when we detect correctly we have not enough space for upload, and we return the correct error response, then on the client side we just see a successful push. | 10:30 |
| valentind | juergbi, Is that intended? Should not we have some warning? | 10:30 |
| tlater[m] | valentind: I'd agree with you that we want a warning. | 10:31 |
| *** slaf_ has joined #buildstream | 10:57 | |
| *** slaf_ has joined #buildstream | 10:57 | |
| *** slaf has quit IRC | 10:59 | |
| *** slaf_ is now known as slaf | 10:59 | |
| *** bochecha has quit IRC | 11:01 | |
| *** abderrahim has joined #buildstream | 11:04 | |
| tiagogomes | I disabled the redirect from buildstream.gitlab.io to buildstream.gitlab.io/buildstream. It looked bad showing a mostly blank page and then finally redirecting to the docs page, and that way we can archive https://gitlab.com/BuildStream/buildstream.gitlab.io/ | 11:13 |
| tiagogomes | But I can put it back if people don't like the idea | 11:14 |
| *** catonano has joined #buildstream | 11:22 | |
| tlater[m] | tiagogomes: I mostly liked the redirect because it's easier to type out buildstream.gitlab.io, and the link looks nicer | 11:56 |
| tlater[m] | Maybe we can make the redirect look like its loading the page instead? i.e., load most of the theme, just no content, so that it transitions more smoothly? | 11:57 |
| tiagogomes | The link would be sorter indeed, but that flashy white page looks horrible | 11:57 |
| tlater[m] | Hrm. Is there any way to just put the front documentation page there? | 11:58 |
| tiagogomes | No :( | 11:58 |
| gitlab-br-bot | buildstream: merge request (mac_fixes->master: Implement compatibility fixes for MacOSX and WSL Blocks #411 and #412") #726 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/726 | 11:58 |
| tlater[m] | Surely if we can put a redirect page on there we could place a copy of the doc page. Alternatively maybe do it with an iframe? | 11:59 |
| tiagogomes | I don't think those two solutions are any better | 12:02 |
| tiagogomes | If we have a machine with a docs.buildtream.build domain, we could instead rsync the documentation to there as part of the CI | 12:03 |
| tlater[m] | tiagogomes: Copying obviously has code duplication issues, an iframe wouldn't load on all browsers... Yeah, this is annoying |: | 12:04 |
| tiagogomes | Actually, we don't need a separate machine | 12:05 |
| tlater[m] | Yes, a subdomain with a reverse proxy on that machine would be enough. | 12:06 |
| tiagogomes | We could just add a CNAME record poiting docs.buidstream.build to buildstream.gitlab.io/buildstream | 12:06 |
| tlater[m] | tiagogomes: Does sphinx link to docs.buildstream.build/page though? | 12:07 |
| tlater[m] | I suppose it would | 12:07 |
| tlater[m] | Time we started using our domain for documentation anyway :) | 12:08 |
| *** iker has quit IRC | 12:14 | |
| *** iker has joined #buildstream | 12:15 | |
| tiagogomes | yes :) | 12:18 |
| tiagogomes | Though we would have to regenerate a new let's encrypt certificate, which is annoying | 12:18 |
| tiagogomes | toscalix are you happy that we use docs.buildstream.build for the documentation? | 12:21 |
| *** tiagogomes has quit IRC | 12:28 | |
| *** tiagogomes has joined #buildstream | 12:28 | |
| tlater[m] | tiagogomes: Doesn't let's encrypt support wildcard certificates these days? | 12:33 |
| tlater[m] | So we could get a certificate for *.buildstream.build, capturing www and anything else we want in the process? | 12:33 |
| * tlater[m] knows this was a requested feature two years ago, that they said they'd be working on | 12:34 | |
| *** tpollard has joined #buildstream | 12:37 | |
| *** lachlan has joined #buildstream | 12:51 | |
| tiagogomes | Perhaps, they have. But in any case, we still would have to request a new one with the wildcard] | 12:53 |
| gitlab-br-bot | buildstream: merge request (Qinusty/message-helpers->master: WIP: Continued work on improving BuildStream messaging API) #670 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/670 | 12:56 |
| *** iker has quit IRC | 13:08 | |
| *** iker has joined #buildstream | 13:09 | |
| *** raoul has joined #buildstream | 13:11 | |
| *** finn has joined #buildstream | 13:12 | |
| finn | Anyone having problems with the docker plugin? | 13:13 |
| finn | It's failing on the fetch for me. | 13:13 |
| finn | I get this: | 13:13 |
| finn | [--:--:--] START docking.bst-0: Fetching image jacobcallahan/auto-tools-docker:latest with digest sha256:bb8040e5b24f5f3c84d7a2b80bc37834a59beaad9832393e71fbce735a4e7c75 | 13:13 |
| finn | [00:00:02] FAILURE docking.bst-0: Fetching image jacobcallahan/auto-tools-docker:latest with digest sha256:bb8040e5b24f5f3c84d7a2b80bc37834a59beaad9832393e71fbce735a4e7c75 | 13:13 |
| finn | [00:00:02] FAILURE docking.bst: 403 Client Error: Forbidden for url: https://production.cloudflare.docker.com/registry-v2/docker/registry/v2/blobs/sha256/b0/b0afedb6056e993a048606918d78c2f9686c231bb9f33478e79d2540022df9db/data?verify=1537970417-DOBplMsSgsWQLHy7Dp5lxytd7b8%3D | 13:13 |
| *** raoul has quit IRC | 13:14 | |
| cs-shadow | finn: a potential problem can be that the image has been updated but the bst file still has an old ref. Do you still get this error after tracking the element? | 13:15 |
| finn | I've tried updating the ref manually and using bst build base.bst --track-all | 13:17 |
| finn | I will log an issue later but it crashes out pretty bad when the fetch fails. I have to kill the process manually | 13:18 |
| cs-shadow | That’s bad. I’m on phone at the moment but I’ll try to reproduce this once I am near my laptop | 13:19 |
| finn | Are there any problems with cloudfare and proxies? | 13:20 |
| cs-shadow | That has happened in the past | 13:20 |
| cs-shadow | And the “fix” for that was just to try again :/ | 13:20 |
| finn | Should I be able to access: https://production.cloudflare.docker.com/ in my browser? | 13:20 |
| finn | I get: {"status":403,"message":"Error: invalid URL signature"} | 13:20 |
| gitlab-br-bot | buildstream: merge request (Qinusty/message-helpers->master: WIP: Continued work on improving BuildStream messaging API) #670 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/670 | 13:22 |
| tlater[m] | finn: Last time this happened we saw that error too | 13:24 |
| tlater[m] | It's probably a hiccup on docker's end | 13:24 |
| finn | that fixes itself? | 13:25 |
| finn | Are you able to fetch a docker image via a plugin atm? | 13:26 |
| *** tpollard has quit IRC | 13:26 | |
| *** raoul has joined #buildstream | 13:27 | |
| tlater[m] | finn: Let me try - last time it took ~30 minutes, but did fix itself. | 13:27 |
| finn | I've had this probelm for ~4h | 13:28 |
| *** raoul has quit IRC | 13:31 | |
| tlater[m] | Ugh, I was in the middle of a rebase so my buildstream's broken... | 13:32 |
| tlater[m] | Actually, I guess I can just run master | 13:33 |
| finn | plz do | 13:33 |
| finn | If you crash, I can rest easy | 13:33 |
| *** tpollard has joined #buildstream | 13:33 | |
| tlater[m] | finn: tracking works for me, fetching seems to do *something* | 13:34 |
| tlater[m] | Might take a few minutes before I can confirm that that something is actually downloading an image, but it looks good. | 13:34 |
| tlater[m] | finn: I can successfully fetch our image-authoring docker image as a base element | 13:36 |
| tlater[m] | Can you share details about the element you're trying to build? | 13:36 |
| *** tpollard has quit IRC | 13:36 | |
| finn | tlater[m], I've just tried this: | 13:41 |
| finn | image: jacobcallahan/auto-tools-docker | 13:41 |
| finn | track: latest | 13:41 |
| finn | ref: bb8040e5b24f5f3c84d7a2b80bc37834a59beaad9832393e71fbce735a4e7c75 | 13:41 |
| finn | I'm now getting: FAILURE 403 Client Error: Forbidden for url: | 13:42 |
| tlater[m] | Same on this end | 13:44 |
| finn | hmm | 13:44 |
| finn | also, is bst crashing bad? | 13:44 |
| tlater[m] | You may need special permissions to load that image - I'm not sure our plugin supports that currently. | 13:44 |
| tlater[m] | finn: Yes, that too. Certainly worth a bug report. | 13:45 |
| finn | raised on external | 13:45 |
| tlater[m] | Yes, looks like we're missing some protection against rogue plugins. | 13:46 |
| finn | Want me to raise on bst too? | 13:47 |
| tlater[m] | Definitely, we want to make buildstream stable even if you use poorly programmed plugins | 13:47 |
| gitlab-br-bot | buildstream: merge request (juerg/platform->master: Make platform independent of context) #829 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/829 | 13:47 |
| *** raoul has joined #buildstream | 13:50 | |
| *** raoul has quit IRC | 14:01 | |
| gitlab-br-bot | buildstream: merge request (Qinusty/message-helpers->master: WIP: Continued work on improving BuildStream messaging API) #670 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/670 | 14:02 |
| *** finn has quit IRC | 14:31 | |
| valentind | Getting the exact same versions of python packages as the builders. And still I get pylint errors. | 14:38 |
| tiagogomes | :( | 14:41 |
| tiagogomes | This https://mail.gnome.org/archives/buildstream-notifications-list/ is way too noisy | 14:41 |
| tiagogomes | Nobody can seriously subscribe to this | 14:41 |
| *** tpollard has joined #buildstream | 14:45 | |
| *** tpollard has quit IRC | 14:45 | |
| toscalix | tlater[m]: you closed #429 but I see an open MR associated to it. Is that MR going to be closed? | 14:59 |
| toscalix | I would prefer to have that MR associated to a ticket if possible | 15:00 |
| *** tpollard has joined #buildstream | 15:03 | |
| tiagogomes | tlater[m] http://docs.buildstream.build/ | 15:06 |
| toscalix | 404 | 15:06 |
| toscalix | https also 404 tiagogomes | 15:06 |
| tiagogomes | this link http://docs.buildstream.build/ ? | 15:07 |
| *** finn has joined #buildstream | 15:08 | |
| tiagogomes | I fear your ISP is caching DNS queries and new entries hasn't been updated yet | 15:09 |
| toscalix | will try a little later then | 15:10 |
| *** tpollard has quit IRC | 15:11 | |
| *** finn has quit IRC | 15:15 | |
| *** alatiera_ has joined #buildstream | 15:23 | |
| *** bochecha has joined #buildstream | 15:34 | |
| *** bochecha has left #buildstream | 15:35 | |
| *** bochecha has joined #buildstream | 15:36 | |
| tlater[m] | tiagogomes: I can see the links, looking good :) | 15:48 |
| tlater[m] | toscalix: that merge request was only affected by #429 | 15:50 |
| tlater[m] | It didn't intend to fix it | 15:50 |
| adds68 | valentind, hey this is the branch which is having issues with project refs: https://gitlab.gnome.org/adds68/giggle/tree/buildstream/buildStream | 15:50 |
| tlater[m] | I'm not sure which issue tracks it, just a second... | 15:50 |
| tlater[m] | toscalix: Ah, this one: https://gitlab.com/BuildStream/bst-external/issues/14 | 15:51 |
| *** toscalix has quit IRC | 15:53 | |
| valentind | adds68, OK, I will have a look at it tomorrow. | 16:00 |
| *** iker has quit IRC | 16:00 | |
| *** lachlan has quit IRC | 16:04 | |
| gitlab-br-bot | buildstream: merge request (valentindavid/cache_server_fill_up->master: WIP: Fix cleanup of cache in server when disk is full) #830 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/830 | 16:27 |
| gitlab-br-bot | buildstream: merge request (valentindavid/cache_server_fill_up->master: WIP: Fix cleanup of cache in server when disk is full) #830 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/830 | 16:29 |
| gitlab-br-bot | buildstream: merge request (valentindavid/cache_server_fill_up->master: WIP: Fix cleanup of cache in server when disk is full) #830 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/830 | 16:29 |
| valentind | tlater[m], juergbi, I have created MR !830 for the fixes for #609 (cache server fills up). Would be nice to have some feedback on it. Maybe tlater[m] you can test it if you have already tests available. | 16:30 |
| valentind | I would like to maybe ask adds68 to deploy it for freedesktop-sdk to see if that really fix the bug in production. | 16:31 |
| adds68 | valentind, i have to configure a new cache soon, i will try it then | 16:31 |
| valentind | adds68, I have branch for 1.2. There is a link in the MR. | 16:32 |
| valentind | https://gitlab.com/BuildStream/buildstream/tree/valentindavid/cache_server_fill_up-1.2 | 16:32 |
| valentind | adds68, I have not reproduced exactly the same bug. But I did see that cleaning did not work correctly. And now at least for my tests, it works. | 16:34 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/acme-challenge->master: ci: add ACME challenge) #831 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/831 | 16:34 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/acme-challenge->master: ci: add ACME challenge) #831 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/831 | 16:36 |
| *** catonano has quit IRC | 16:36 | |
| *** finn has joined #buildstream | 17:29 | |
| *** finn has quit IRC | 17:41 | |
| *** toscalix has joined #buildstream | 18:10 | |
| *** toscalix has quit IRC | 18:10 | |
| *** toscalix has joined #buildstream | 18:12 | |
| *** toscalix has quit IRC | 18:14 | |
| *** catonano has joined #buildstream | 18:44 | |
| *** catonano has quit IRC | 19:50 | |
| gitlab-br-bot | buildstream: merge request (chandan/bst-checkout-build->master: Add `--deps build` option to `bst checkout`) #819 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/819 | 20:14 |
| *** catonano has joined #buildstream | 20:16 | |
| *** catonano has quit IRC | 21:16 | |
| *** bochecha has quit IRC | 22:20 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!