| *** toscalix has quit IRC | 02:09 | |
| *** ctbruce has joined #trustable | 08:38 | |
| *** leeming has quit IRC | 10:42 | |
| *** tiagogomes has quit IRC | 10:42 | |
| *** ctbruce has quit IRC | 10:42 | |
| *** sambishop has quit IRC | 10:42 | |
| *** ctbruce has joined #trustable | 10:54 | |
| *** tiagogomes has joined #trustable | 10:55 | |
| *** leeming has joined #trustable | 10:55 | |
| *** sambishop has joined #trustable | 10:55 | |
| *** leeming has quit IRC | 11:33 | |
| *** leeming has joined #trustable | 11:33 | |
| paulsherwood | Kinnison: any suggestions on least-effort approach to establish a test framework for future content in https://mustard.trustable.io/trustable-mustard/HEAD/ ? | 13:49 |
|---|---|---|
| *** toscalix has joined #trustable | 14:34 | |
| *** toscalix has quit IRC | 14:42 | |
| *** toscalix has joined #trustable | 14:45 | |
| *** toscalix has quit IRC | 14:52 | |
| *** AlisonChaiken has quit IRC | 15:12 | |
| Kinnison | testing changes to a mustard? Interesting idea | 15:32 |
| * Kinnison would have to ponder | 15:32 | |
| *** toscalix has joined #trustable | 15:49 | |
| paulsherwood | i'm thinking about the meta-problem, in a way... but on a call. will explain soon | 15:49 |
| Kinnison | Sure, if you burble here, I'll see it later :-D | 15:52 |
| *** toscalix has quit IRC | 15:55 | |
| *** toscalix has joined #trustable | 16:08 | |
| *** toscalix has quit IRC | 16:11 | |
| *** toscalix has joined #trustable | 16:14 | |
| *** Gravious has joined #trustable | 16:14 | |
| *** toscalix has quit IRC | 16:15 | |
| paulsherwood | so i'm imagining that https://mustard.trustable.io/trustable-mustard becomes a kind of in-miniature representation of a trustable project flow in itself | 16:16 |
| paulsherwood | i.e. we clarify requirements (for the trustable) there... and then create tests to represent validation of the logic in the requirements | 16:17 |
| paulsherwood | ideally the process that we use to operate on the mustard would itself conform to what we know to be required... so there'd be a need for identity, code review, ci, cd etc | 16:18 |
| *** toscalix has joined #trustable | 16:18 | |
| Kinnison | I'm not entirely sure how to write tests which validate that there exists software or workflows to satisfy the requirements of trustability | 16:18 |
| * Kinnison is still fairly new to thinking about all this stuff and hasn't onboarded everything in the project yet | 16:19 | |
| paulsherwood | Kinnison: i don't have a thorough answer. but there are some simple partial answers... | 16:20 |
| paulsherwood | - we can confirm valid yaml | 16:20 |
| paulsherwood | - for fields we decide need to be given-when-then, we check that | 16:21 |
| paulsherwood | - maybe there are other meta-tests we can start with, until a brainwave hits :) | 16:21 |
| paulsherwood | this is just a straw man after all | 16:22 |
| *** toscalix has quit IRC | 16:23 | |
| Kinnison | That'd be a way to syntax-check the mustard content, but it wouldn't be any kind of semantic validation | 16:23 |
| *** ctbruce has quit IRC | 16:24 | |
| paulsherwood | i agree... but it's a start? | 16:25 |
| Kinnison | Perhaps. Though shouldn't that really be a property of Mustard itself (to lint its input) ? | 16:25 |
| Kinnison | Linting the yaml is the major feature of Myrosinase :-) | 16:26 |
| Kinnison | THough it doesn't understand loom :-) | 16:26 |
| paulsherwood | what is that? | 16:26 |
| Kinnison | Myrosinase is an incomplete alternative implementation of Mustard | 16:26 |
| *** AlisonChaiken has joined #trustable | 16:26 | |
| Kinnison | I wrote it years ago | 16:26 |
| Kinnison | when i was learning haskell | 16:26 |
| paulsherwood | oh. i believe there's also a jpohlmann incomplete implementation in closure | 16:27 |
| Kinnison | Quite possibly | 16:28 |
| Kinnison | I started to investigate the complexity of good yaml parsing with location information in Rust | 16:28 |
| paulsherwood | in any case i'm using these as example tests | 16:28 |
| Kinnison | Nod, linting the mustard is a good start | 16:28 |
| paulsherwood | building them into the tooling defeats my (strange, little) purpose | 16:28 |
| Kinnison | Hmm | 16:29 |
| paulsherwood | my reasoning is that afaik i there is no model of the workflow we're describing, anywhere in the open | 16:31 |
| Kinnison | Of the whole workflow, absolutely not, it's novel. | 16:31 |
| paulsherwood | so i'm just trying to cause one to exist, as a means of further thinking about the workflow itself | 16:31 |
| paulsherwood | (and also helping to tighten up the logic/text we've been discussing) | 16:32 |
| Kinnison | nod. | 16:32 |
| paulsherwood | incidentally https://gitlab.com/trustable/trustable-mustard/issues/2 | 16:36 |
| Kinnison | thanks for capturing that | 16:37 |
| paulsherwood | :) | 16:37 |
| *** toscalix has joined #trustable | 16:38 | |
| *** toscalix has quit IRC | 16:45 | |
| AlisonChaiken | I got a long email over the weekend from someone you likely know, Kinnison. | 16:47 |
| AlisonChaiken | His name is Jamey Sharp, and he's written a tool called Corrode to automatically translate C to Rust. | 16:47 |
| * Kinnison is aware of that name, but hasn't met the person | 16:47 | |
| AlisonChaiken | Corrode is apparently written in Haskell. He's looking for automotive support, so I told him to inquire on trustable mailing list. | 16:48 |
| * paulsherwood wonders how feasible that translation is for realistic established projects | 16:48 | |
| AlisonChaiken | I know. But I hate to discourage people. | 16:48 |
| paulsherwood | :) | 16:48 |
| AlisonChaiken | It's a good idea! | 16:48 |
| AlisonChaiken | But still. | 16:48 |
| paulsherwood | i agree it's a good idea | 16:48 |
| * AlisonChaiken returns to wrasslin' the build system. | 16:48 | |
| paulsherwood | which build system is that, this time? :) | 16:49 |
| AlisonChaiken | Remarkably, we are running Bazel, the open-source version of Google's internal build system. | 17:15 |
| paulsherwood | ooh... how is it? | 17:16 |
| AlisonChaiken | Because we build our rootfs in the same system that we build our cloud applications, so that we can test them together. | 17:16 |
| AlisonChaiken | Like other build systems, it's intricate and painful. I'm trying to get swupdate to build, and just kill me now and put me out of my misery. | 17:16 |
| paulsherwood | rofl | 17:17 |
| paulsherwood | i'll quote you :) | 17:17 |
| AlisonChaiken | This fall I got Qt-GStreamer to build in it, for which I deserve a medal. | 17:17 |
| * Kinnison isn't sure quoting her will resolve AlisonChaiken's pain | 17:17 | |
| * paulsherwood wishes he could be at https://openiotelcna2017.sched.com/event/9Ith/tutorial-building-the-simplest-possible-linux-system-rob-landley-se-instrumentscom | 17:18 | |
| AlisonChaiken | paulsherwood, ahem, have you heard Landley speak before? | 17:23 |
| * Kinnison wonders if it'll be video'd | 17:23 | |
| paulsherwood | AlisonChaiken: i've seen him on video | 17:24 |
| AlisonChaiken | I've tried listening to his talks in person twice, but he tends to rant about how much he hates the GPL and suchlike. | 17:25 |
| AlisonChaiken | However you feel about the GPL, I prefer talks that stick to the announced topic. | 17:25 |
| paulsherwood | ack | 17:26 |
| * Kinnison 's talks tend to devolve into rants about github :-D | 17:28 | |
| Kinnison | At least, the ones about my git server software do :-P | 17:28 |
| *** toscalix has joined #trustable | 18:10 | |
| *** AlisonChaiken has quit IRC | 18:41 | |
| *** AlisonChaiken has joined #trustable | 18:59 | |
| *** Gravious has quit IRC | 19:02 | |
| *** toscalix has quit IRC | 19:16 | |
| *** toscalix has joined #trustable | 20:36 | |
| *** toscalix has quit IRC | 21:28 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!