| *** alatiera has joined #buildstream | 01:32 | |
| *** alatiera has quit IRC | 01:33 | |
| *** bochecha has quit IRC | 01:54 | |
| *** max has quit IRC | 01:56 | |
| *** xjuan has quit IRC | 01:57 | |
| *** nimish has joined #buildstream | 02:31 | |
| *** nimish has quit IRC | 03:23 | |
| *** werew has joined #buildstream | 04:23 | |
| *** kapil___ has joined #buildstream | 04:27 | |
| *** nssy has joined #buildstream | 04:34 | |
| *** ManiacTwister has joined #buildstream | 06:22 | |
| *** alatiera has joined #buildstream | 07:02 | |
| *** Kinnison has quit IRC | 07:32 | |
| *** Kinnison has joined #buildstream | 07:45 | |
| *** kapil___ has quit IRC | 08:17 | |
| *** slaf_ has joined #buildstream | 08:50 | |
| *** slaf has quit IRC | 08:52 | |
| *** slaf_ is now known as slaf | 08:52 | |
| gitlab-br-bot | coldtom opened issue #825 (Stack Trace while trying to use autocompletions on `bst workspace open` with mistyped initial segment) on buildstream https://gitlab.com/BuildStream/buildstream/issues/825 | 08:59 |
|---|---|---|
| *** toscalix has joined #buildstream | 09:30 | |
| jennis | adds68, there is not currently, although the information is there as we use this to update the file recording the cache size iirc | 09:47 |
| jennis | adds68, were you wondering whether you could check the size *before* you `bst pull`...? Or just know the size once you've `bst pull`ed? | 09:48 |
| adds68 | jennis, ah ok, currently i can never tell if my network is super slow or bst has stopped pulling from the cache | 09:49 |
| adds68 | It would be nice instead of time, to report xMb of YMb downloaded | 09:50 |
| jennis | Or potentially something we can work into the pull log | 09:55 |
| jennis | adds68, sounds like you want to know whther bst is pulling or not, regardless of the size...? | 09:56 |
| adds68 | jennis, well it has the timer, but llvm reported 4 hours the other day, which is not correct, as it usually takes ~20 minutes | 09:57 |
| adds68 | So i don't trust it | 09:58 |
| benschubert | jjardon: do we only have one aarch64 runner for buildstream-docker-images? https://gitlab.com/BuildStream/buildstream-docker-images/pipelines/40687623 seems to have only one job running | 10:10 |
| *** bochecha has joined #buildstream | 10:58 | |
| toscalix | valentind: around? | 11:09 |
| valentind | yes | 11:09 |
| toscalix | I thi Sander's request to provide to the overall community background about the infra is a good one | 11:10 |
| toscalix | think | 11:10 |
| *** max has joined #buildstream | 11:10 | |
| toscalix | It would good to describe the servers we are using, what for and who is in charge | 11:11 |
| jmac | I'm not sure what you mean by infrastructure in that email. Haven't we always had one? | 11:11 |
| toscalix | and some background on who was providing them earlier | 11:11 |
| toscalix | jmac: yes, but I guess we haven been explicit about it | 11:12 |
| jmac | No, that's true | 11:12 |
| toscalix | so telling everybody what we are using and who is in change is a good thing | 11:12 |
| bochecha | just listing the services would have done a lot of difference :) | 11:12 |
| *** milloni has quit IRC | 11:13 | |
| toscalix | we should take this opportunity to find a second maintainer of the infra, since there are vacations, sick leaves etc. | 11:13 |
| toscalix | valentind: is thi soemthing you can do? | 11:14 |
| bochecha | also, the buildstream package is now in Fedora 30, as well as in the updates-testing repository for Fedora 29 and 28 :) | 11:14 |
| toscalix | send a mail to the mailing list informing about this? | 11:14 |
| valentind | OK | 11:15 |
| toscalix | bochecha: are you suggesting to use this new subgroup to store those? | 11:15 |
| valentind | But it is so minor, I wonder why we need to have a discussion about it. | 11:15 |
| valentind | I have a repository with the ansible configuration for the sleds. | 11:15 |
| toscalix | it is not about the info itself, it is about transparency | 11:16 |
| valentind | Next time I have to change something, I will make more ansible configuration. | 11:16 |
| bochecha | toscalix: to store what? | 11:16 |
| toscalix | somebody is in chanrge of a key area for the project and people wants to know details about it. I see it as a normal thing | 11:16 |
| toscalix | bochecha: nevermind | 11:17 |
| *** milloni has joined #buildstream | 11:17 | |
| jjardon | benschubert: you need to ask valentind | 11:17 |
| valentind | benschubert, we have 8 builders. | 11:17 |
| toscalix | bochecha: it would be a good thing to announce the package availability on the mailing list and include the links in the web | 11:18 |
| *** kapil___ has joined #buildstream | 11:18 | |
| benschubert | jjardon: thanks :) | 11:18 |
| benschubert | valentind oh... I only got one job at a time, is there a way of seeing their health / what they are doing somewhere? | 11:18 |
| tpollard | juergbi: I'm asking in relation to 'buildstream_pb2.GetReferenceRequest()' | 11:18 |
| bochecha | toscalix: I was planning on doing that once they hit stable | 11:18 |
| bochecha | toscalix: as well as send a MR to change the install instructions | 11:18 |
| tpollard | juergbi: in pull for instance, we assign the ref to it's key field | 11:19 |
| juergbi | tpollard: this is a gRPC generated class based on the .proto | 11:19 |
| toscalix | bochecha: fair enough | 11:19 |
| tpollard | I'm looking to pass another field with it | 11:19 |
| juergbi | that would be a proto extension | 11:19 |
| valentind | benschubert, you are right. They seem to be disabled. | 11:19 |
| valentind | benschubert, they got disconnected from the network. | 11:26 |
| benschubert | valentind: ouch. Would it be possible to have healthchecks somewhere? not saying now, but as a long term goal :) | 11:27 |
| valentind | benschubert, like this? https://gitlab.com/groups/BuildStream/-/settings/ci_cd | 11:32 |
| benschubert | valentind: ah, I get a 404, guess I don't have enough rights :/ | 11:33 |
| valentind | Probably | 11:33 |
| valentind | sled22 is not working at all. | 11:33 |
| valentind | Ah, it is booting. | 11:35 |
| valentind | I am redeploying. It might take 15-30 minutes. | 11:35 |
| benschubert | thanks a lot! | 11:35 |
| tpollard | juergbi: basically I'm trying to let GetReference handle artifacts with missing buildtrees | 11:38 |
| juergbi | tpollard: how do you want to extend the request for this? | 11:39 |
| juergbi | the ReferenceStorage service doesn't know anything about the artifact structure. and it shouldn't, imo | 11:39 |
| tpollard | internally it calls _reachable_refs_dir which falls over in this case, so I'm looking to have an argument to allow it to skip it | 11:39 |
| juergbi | you want to completely skip it or restrict it to certain subdirs? | 11:40 |
| tpollard | handle it as with pull, i'e if dirnode.name in 'excluded_dirs' or whatever, then don't bother walking into it | 11:40 |
| juergbi | I'm wondering whether we should turn it around | 11:41 |
| tpollard | I've got it working clientside, but within casserver I have no way to reference back to the context of the element that is being requested | 11:41 |
| juergbi | i.e., optionally specify which subdirs we're interested in and then potentially even return the tree digests for each subdir | 11:41 |
| juergbi | if you add a field to the request in the .proto file, and then regenerate the code, you should be able to access the field also on the server side | 11:42 |
| juergbi | i.e., same way as key works right now | 11:43 |
| tpollard | 'message GetReferenceRequest' in buildstream.proto ? | 11:45 |
| juergbi | yes | 11:46 |
| juergbi | check CONTRIBUTING.rst for how to regenerate the code | 11:48 |
| gitlab-br-bot | jjardon opened (was WIP) MR !1014 (jjardon/license_badge->master: README.rst: Add license badge) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1014 | 11:51 |
| jmac | Is there a quick fix for the x86-64/x86_64 issue that's stopping freedesktop-sdk building? | 11:59 |
| valentind | jmac, you mean a work around? | 12:00 |
| jmac | Yes, I just want to build it to test something else | 12:00 |
| valentind | You can set -o bootstrap_build_arch x86_64 -o target_arch x86_64 | 12:00 |
| valentind | betwenn "bst" and "build" in the command line. | 12:00 |
| jmac | That works, many thanks | 12:01 |
| jmac | (I altered ARCH_OPTS in Makefile with the same options) | 12:02 |
| *** raoul has joined #buildstream | 12:10 | |
| jmac | Actually, when I tried that directly, it says "Unknown option 'bootstrap_build_arch' specified on the command line" | 12:24 |
| jmac | Wait, I see, that's something specific to freedesktop-sdk | 12:26 |
| *** nimish has joined #buildstream | 12:26 | |
| *** nimish has quit IRC | 12:36 | |
| bochecha | jmac: yeah, the -o options will usually be per project (defined in project.conf) | 12:37 |
| *** nimish has joined #buildstream | 12:37 | |
| *** raoul has quit IRC | 12:46 | |
| *** nimish has quit IRC | 12:52 | |
| *** nimish has joined #buildstream | 12:52 | |
| gitlab-br-bot | aevri closed MR !957 (aevri/safe_noninteractive->master: BREAK: make destructive action scripts consistent) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/957 | 12:55 |
| *** lsfranco has quit IRC | 13:07 | |
| *** lsfranco has joined #buildstream | 13:08 | |
| *** max has quit IRC | 13:11 | |
| *** lsfranco has quit IRC | 13:18 | |
| *** lsfranco has joined #buildstream | 13:18 | |
| gitlab-br-bot | aevri opened issue #826 (Remove auto-init behaviour) on buildstream https://gitlab.com/BuildStream/buildstream/issues/826 | 13:52 |
| *** nimish has quit IRC | 14:07 | |
| *** nimish has joined #buildstream | 14:07 | |
| *** lachlan has joined #buildstream | 14:16 | |
| *** abderrahim has quit IRC | 14:19 | |
| *** abderrahim has joined #buildstream | 14:19 | |
| gitlab-br-bot | cs-shadow opened issue #828 (Sort NEWS before release) on buildstream https://gitlab.com/BuildStream/buildstream/issues/828 | 14:29 |
| *** nimish has quit IRC | 14:30 | |
| *** nimish has joined #buildstream | 14:33 | |
| gitlab-br-bot | willsalmon closed issue #740 (`bst shell --build` should let the user choose whether to use cached build trees) on buildstream https://gitlab.com/BuildStream/buildstream/issues/740 | 14:35 |
| gitlab-br-bot | willsalmon merged MR !986 (willsalmon/shellBuildTrees->master: Shell --build has optional buildtrees) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/986 | 14:35 |
| tpollard | juergbi: what would you expect the behaviour to be for the user if they specific --pull-buildtrees, but the remote artifact is missing the blob for it? | 14:45 |
| tpollard | We already capture the missing blob exception, just debating what the best course of action is at a user level | 14:45 |
| juergbi | tpollard: I don't think this should cause any fatal errors | 14:46 |
| juergbi | a warning could make sense | 14:46 |
| juergbi | or maybe even just an info message | 14:47 |
| *** nimish has quit IRC | 14:47 | |
| *** nimish has joined #buildstream | 14:48 | |
| gitlab-br-bot | willsalmon opened issue #829 (Download buildtrees on damand for bst shell --build) on buildstream https://gitlab.com/BuildStream/buildstream/issues/829 | 14:50 |
| tpollard | I think if they've explicitly asked for the buildtree then we shouldn't pull the partial artifact. maybe this could come under configurable warnings I'm not sure | 14:50 |
| tpollard | but I'm not 100% sure | 14:51 |
| tpollard | hmm | 14:51 |
| valentind | When building with buildgrid and buildbox, at some point I get error StatusCode.RESOURCE_EXHAUSTED from buildbox. And then it stops working. | 14:54 |
| *** nimish has quit IRC | 15:03 | |
| *** nimish has joined #buildstream | 15:04 | |
| juergbi | tpollard: my point of view is from pull-buildtrees in userconfig | 15:04 |
| juergbi | if it's set globally, we shouldn't fail | 15:04 |
| juergbi | as otherwise it's not usable as global option | 15:04 |
| juergbi | and different behavior between userconfig and CLI option would surprise me as a user | 15:05 |
| juergbi | if you run a command where you specify to actually use the buildtree, e.g., bst shell --use-buildtree=... would fail in any case if there is no buildtree, no matter whether the artifact is local-only or pulled from a remote | 15:06 |
| *** nimish has quit IRC | 15:07 | |
| tpollard | juergbi: would you find it acceptable if _fetch_directory utilised _reachable_refs_dir for generating the blobs it needs to fetch, instead of having it's own loop? | 15:25 |
| *** nimish has joined #buildstream | 15:30 | |
| gitlab-br-bot | LaurenceUrhegyi approved MR !1014 (jjardon/license_badge->master: README.rst: Add license badge) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1014 | 15:35 |
| *** nimish has quit IRC | 15:45 | |
| *** lsfranco has quit IRC | 15:48 | |
| *** lsfranco has joined #buildstream | 15:49 | |
| *** lachlan has quit IRC | 15:50 | |
| *** max has joined #buildstream | 16:05 | |
| *** bochecha has quit IRC | 16:09 | |
| gitlab-br-bot | jjardon merged MR !1014 (jjardon/license_badge->master: README.rst: Add license badge) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1014 | 16:09 |
| *** lsfranco has quit IRC | 16:19 | |
| *** lsfranco has joined #buildstream | 16:19 | |
| *** lsfranco has quit IRC | 16:34 | |
| *** lsfranco has joined #buildstream | 16:34 | |
| *** max has quit IRC | 16:58 | |
| gitlab-br-bot | willsalmon opened (was WIP) MR !1017 (716-add-example-with-build-directory-outside-of-source-directory-2->willsalmon/outOfSourecBuilddocs: Add example with build directory outside of source directory) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1017 | 17:04 |
| gitlab-br-bot | willsalmon opened MR !1018 (willsalmon/outOfSourecBuilddocs->master: Adding Out of Source Build Example) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1018 | 17:09 |
| gitlab-br-bot | willsalmon closed MR !1017 (716-add-example-with-build-directory-outside-of-source-directory-2->willsalmon/outOfSourecBuilddocs: Add example with build directory outside of source directory) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1017 | 17:09 |
| gitlab-br-bot | aevri opened (was WIP) MR !1015 (aevri/rm-autoinit->master: BREAK: remove auto-init behaviour) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1015 | 17:28 |
| gitlab-br-bot | jmacarthur opened issue #830 (_relative_symlink_target depends on the host filing system) on buildstream https://gitlab.com/BuildStream/buildstream/issues/830 | 17:31 |
| jmac | I've spent most of the day trying to figure out the symlinks issue. Hopefully #830 is a reasonable summary of it | 17:40 |
| jjardon | Hi, Is this page meant to be there? https://buildstream.build/detail.html | 17:56 |
| jjardon | toscalix: ^ | 17:56 |
| jjardon | Is there a page somewhere with reasons to use buildstream above other build systems? | 17:57 |
| toscalix | jjardon: no there is not. I think that would be a topic for a post, signed by somebody, instead of something on the web which "represents everybody" | 18:09 |
| toscalix | that would be a corporate approach | 18:09 |
| toscalix | this is my view on doing benchmarking or feature comparassions with other open source projects | 18:09 |
| toscalix | it is not elegant in general | 18:10 |
| toscalix | I do not see it as a competition | 18:10 |
| jjardon | toscalix: I do not understand what you mean by corporate, I'm only interested in the technical arguments | 18:10 |
| jjardon | I think https://docs.buildstream.build/main_about.html#why-should-i-use-buildstream is a good start | 18:10 |
| toscalix | even technical arguments can be arguable | 18:10 |
| *** drakythe has joined #buildstream | 18:10 | |
| toscalix | agree | 18:11 |
| toscalix | talking about your strengths is the right approach | 18:11 |
| jjardon | Can someone review https://gitlab.com/BuildStream/website/merge_requests/105 , please? | 18:13 |
| toscalix | corporate = commercial product, for instance | 18:14 |
| toscalix | when you position your product "against" the segment leader | 18:14 |
| jjardon | I do not see why is that any different from competition with FOSS projects. Pretty common to have a page with differences/advantages over the others | 18:22 |
| *** toscalix has quit IRC | 18:30 | |
| cs-shadow | jumping late to this convo but I think such a page may also help users answer the question of what is the right tool for the job, and is buildstream the right tool for their use-case | 18:54 |
| cs-shadow | it's not always about pointing out deficiencies in other projects, but more about highlighting similarities and differences as I see it | 18:55 |
| jjardon | exactly | 18:57 |
| *** alatiera has quit IRC | 19:23 | |
| gitlab-br-bot | cs-shadow opened MR !1019 (chandan/abstract-git-source->master: Expose base class for Git source plugins) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1019 | 20:07 |
| *** juergbi has quit IRC | 21:00 | |
| *** juergbi has joined #buildstream | 21:01 | |
| *** kapil___ has quit IRC | 21:17 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!