IRC logs for #trustable for Thursday, 2018-09-06

*** brlogger has joined #trustable01:18
*** brlogger` has quit IRC01:19
*** poppy has joined #trustable07:09
*** poppy is now known as spinglet07:09
*** toscalix has joined #trustable08:36
*** Shaun has joined #trustable09:01
paulsherwoodare we tracking issues/activities anywhere?09:43
flatmushminimal distro gitlab has an issue tracker09:43
paulsherwoodfor a while we were using trello, but I suggest we just go for gitlab in general09:43
flatmushare you wanting us to use the issues like a kanban though?09:44
paulsherwoodideally, it would be nice09:45
* paulsherwood is crap at doing the kanban dance himself, though09:45
flatmushyeah, I don't really like kanban that much, but of the approaches I think trello is best09:46
flatmushissue trackers should be for issues imo09:46
flatmushSTANDUP! Order: flatmush, Shaun, iker09:46
flatmush## Ben Brewer (flatmush)09:46
flatmush* Done09:46
flatmush  - Got memtester building as part of minimal-distro test group09:46
flatmush* Doing09:46
flatmush  - Figure out best way of testing inside QEMU (probably ssh)09:46
flatmush  - Execute memtester as a test using the CI09:46
flatmush## Shaun Mooney (Shaun)09:46
Shaun* Done09:46
Shaun  - Read through relevant parts of "Engineering a Safer World" book09:46
Shaun* Doing09:46
Shaun  - Working through STPA examples09:46
Shaun  - Making notes regarding STPA and linux09:46
Shaun* Next09:46
Shaun  - Create simple Functional Control Model (STPA) for autonomous cars09:46
Shaun  - Summarise STPA material09:46
Shaun    - Explain autonomous car functional control model as an example09:46
Shaun* Issues09:47
Shaun  - None09:47
Shaun## Iker Perez ( iker )09:47
iker## Iker Perez09:47
iker* Done09:47
iker   - Watch BuildStream Fosdem talk09:47
iker   - Have a short BuildStream lesson from jennis09:47
iker   - Create a virtual machine with buildstream installed09:47
iker* Doing09:47
iker   - Follow the BuildStream tutorial09:47
iker* Next09:47
iker    - Analyse the way the minimal distro uses buildstream09:47
iker    - Install QEMU in the virtual machine09:47
iker# Discussion09:47
flatmushSTANDUP ENDS!09:52
paulsherwoodShaun: where will your WIP STPA model reside please?09:53
* paulsherwood is aware that several people are already interested in the work, even though it has barely begun :-)09:54
Shaunpaulsherwood: At the moment on paper as I'm still getting to grips with how the whole process works09:56
ShaunOnce I have something better where is a good place to put it?09:56
ShaunTrustable wiki?09:56
paulsherwoodnot wiki, please let's have a new repo, something like av-stpa09:59
paulsherwoodhow will the model be expressed? just plain text, or markdown, or do we need to diagram for it? (e.g. mermaid, plantuml)10:00
ShaunCombination of block diagrams, tables and plain text10:01
ShaunI was about to ask if anybody has any recommendations for a good format for those ^10:01
paulsherwoodwell, that can all be gitlab markdown + mermaid, should render nicely10:02
paulsherwoodand by having it in a repo, we can consider CI/CD and gitect on the content in due course10:02
Shaunpaulsherwood: Do you have some time for a quick chat? I just want to make sure we're on the same page before I continue10:03
paulsherwoodyup absolutely10:03
paulsherwoodbefore i forget has an example of a pretty mermaid diagram10:03
paulsherwoodtables in markdown are a bit rubbish but hopefully they'll be enough for what we need10:13
Shaunpaulsherwood: mermaid looks like exactly what I wanted :)10:17
flatmushthat approach does seem a lot less crazy11:15
paulsherwoodi wholeheartedly agree11:17
flatmushI've not read far enough, but the only issue I can see is that it looks like this doesn't lead to any failure rate type numbers?11:18
paulsherwoodthey shouldn't be necessary11:18
flatmushwell I don't think they'd be accurate anyway, so it's a moot point, but I suspect at some point in the chain, someone will think they want them (like estimates)11:19
paulsherwoodit's fallacious thinking to assume event 1 (x%) and event 2 (y%) leads to hazard xy%11:19
paulsherwoodi've already read enough to debunk11:19
paulwaters_paulsherwood, Shaun ^^12:37
paulsherwoodShaun: which gitlab id do we use for you in the trustable group?12:38
Shaunpaulsherwood: shaunmooneyct12:38
paulsherwoodok thanks12:38
paulsherwoodpaulwaters_: interesting!12:39
paulwaters_and another....12:41
paulwaters_paulsherwood, Shaun ^^12:41
paulsherwoodeven better :-)12:44
Shaunpaulwaters_: Excellent12:44
paulwaters_think I've peaked with that one12:46
paulwaters_those two were presented discussed at the 2018 PSASS workshop, other papers can be found of the links from their homepage here:
ShaunThis one talks a big game about having done the analysis and the control model on page 9 looks good, but they don't include the results they say they have
ShaunThey used XSTAMPP which I had seen before. the Thomas thesis alludes to programmes which will automate the process for complex systems12:55
paulsherwoodShaun: maybe email the authors, ask for the results?12:58
Shaunpaulsherwood: will have a look around first to see if I can find them, then will email12:59
*** traveltissues has quit IRC13:50
*** traveltissues has joined #trustable13:50
*** traveltissues has quit IRC14:13
*** Shaun has quit IRC14:58
*** spinglet has quit IRC15:37
*** iker has quit IRC16:00
*** toscalix has quit IRC18:23

Generated by 2.15.3 by Marius Gedminas - find it at!