IRC logs for #trustable for Friday, 2018-07-20

*** willbarnard has joined #trustable05:58
*** weyfonk has joined #trustable08:13
*** traveltissues has joined #trustable08:47
paulsherwoodflatmush: is there a cache server involved in your ci? if so, would a re-run be faster?09:02
paulsherwoodalso maybe we can adjust the timeout09:03
paulsherwoodor create a dedicated runner for this09:03
flatmushshould be the same as definitions, so probably09:04
flatmushbut still seems wrong for the CI to timeout in less than the time a full rebuild takes09:05
flatmushnot very automated09:05
flatmushAh, we're not using the cache, I guess I could nag ironfoot for the key09:22
*** willbarnard has quit IRC09:29
*** willbarnard has joined #trustable09:30
ironfootyes you could, but I believe it wasn't even working in baserock/definintions.git so I'm not sure it will speed things up09:32
ironfootlet's give it a try though09:33
flatmushSTANDUP! Order: flatmush, noisecell, willbarnard10:09
flatmush## Ben Brewer (flatmush)10:09
flatmush* Done10:09
flatmush  - Make minimal distro work public10:09
flatmush  - Try to get CI working10:09
flatmush  - Discussion about bootstrapping10:09
flatmush  - Work on minimal distro10:09
flatmush* Doing10:09
flatmush  - Enable CI caching10:09
flatmush  - Discussion about bootstrapping compilers10:09
flatmush  - Attempt to replace glibc with musl10:09
flatmush## Francisco Redondo Marchena (noisecell)10:09
noisecell* Done10:09
noisecell    - Update package dependencies to use build and runtime dependencies wisely10:09
noisecell* Doing10:09
noisecell  - Trim the system and update packages to latest versions.10:09
noisecell    - Update packages to latests stable versions10:09
noisecell## Will Barnard (willbarnard)10:09
willbarnard* Doing10:10
willbarnard  - Analyzing certification requirements for ISO 26262/IEC 6150810:10
willbarnard    - Planning for functional safety workshop10:10
flatmush  - Improving level of detail on standup notes.10:10
flatmush## Discussion10:10
ironfooto/10:10
noisecell_o_10:11
flatmushironfoot?10:11
ironfootregarding musl, there was a previous attempt done in baserock10:11
ironfootand i think a finished one10:11
flatmushyes, by rdale10:11
ironfoot_o_10:11
flatmushI'm looking at that, it doesn't directly work10:11
flatmushprobably did at the time10:11
flatmushSTANDUP ENDS!10:11
*** willbarnard has quit IRC11:27
*** willbarnard has joined #trustable11:27
paulsherwoodhttps://twitter.com/devcurmudgeon/status/102027048839898726411:37
ltuI may send that to the C Safety and Security Study Group11:49
ltuWe had a debate back when it first started out over whether to make the (hopefully) resulting standard free or not11:50
ltuDavid A Wheeler pointed out that the only way to make a standard relevant was to make it free.11:51
paulsherwoodltu: it's about time i wrote to that group, in any case12:28
paulsherwoodbut by all means send it :)12:29
*** noisecell has quit IRC14:22
persiaETSI has a requirement to charge for standards baked in to the structure of the organisation.  One trick that is used by a number of ETSI working groups is to publish a) the source of the standard (for review), b) a set of requirements for implementation of the standard (different language to standard, but ideally same content), and c) an open source reference implementation.14:56
persiaThose groups don't end up seeing much revenue, but they do seem to be leading with adoption, despite the standard costing money.14:56
persia(in that implementing organisations purchase one copy and nobody ever reads that copy: it's just for the audit trail)14:56
* paulsherwood likes the sound of that :)15:00
persiaAt least for (a), there are an increasing number of standards organisations that are increasing transparency and opportunity for public comment, often under an "open source" initiative.  There were a few people from standards organsiations (or organisations providing services to standards organisations) at OSCON this year, although I don't want to speak in advance of anyone's announcement.15:11
persia(b) and (c) require a lot more commitment by members of the relevant working group for the standard in question, rather than being something that can be required by policy.15:12
*** willbarnard has quit IRC15:39
*** willbarnard has joined #trustable15:40
*** willbarnard has quit IRC16:55
*** toscalix has joined #trustable17:03
*** toscalix has quit IRC17:08
*** toscalix has joined #trustable17:13
*** weyfonk has quit IRC17:32
*** toscalix has quit IRC18:34

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!