IRC logs for #trustable for Wednesday, 2018-07-04

*** toscalix has joined #trustable08:08
*** weyfonk has joined #trustable08:11
*** noisecell has joined #trustable09:04
*** willbarnard has quit IRC09:12
*** willbarnard has joined #trustable09:12
paulwaters_standup in 12 minutes09:33
paulwaters_Kinnison, this is where we are holding standups for the Trustable Distros work09:41
KinnisonOoooh interesting09:42
KinnisonMind if I observe?09:42
paulwaters_Kinnison, please do :)09:42
flatmushSTANDUP! Order: flatmush, noisecell, willbarnard09:45
flatmush## Ben Brewer (flatmush)09:45
flatmush* Done09:45
flatmush  - Initial discussions09:45
flatmush  - Reading trustable documentation09:45
flatmush  - Initial research into BuildStream, BuildRoot and mkroot09:45
flatmush* Doing09:45
flatmush  - Further research into BuildStream, etc.09:45
flatmush  - Attempt to build minimal linux image09:45
flatmush  - Work with fran on Aarch64 image building/bootstrapping.09:45
flatmush## Francisco Redondo Marchena (noisecell)09:45
noisecell* Done09:45
noisecell  - Kick off meeting09:45
noisecell* Doing09:45
noisecell  - Investigate aarch64 bootstrap using buildstream09:45
noisecell    - So far I've cross-built stage2 and get the source bundle for stage309:45
noisecell      but when trying to build natively stage3 inside of stage2 chroot in the09:45
noisecell      board the component build scripts fail.09:45
noisecell      Looking of how these scripts are generated (they are templates09:45
noisecell      which are filled with commands on buildstream build time)09:45
noisecell      For reference, Im using the instructions given in:09:45
*** toscalix has quit IRC09:45
flatmush## Will Barnard (willbarnard)09:46
willbarnard* Doing09:46
willbarnard  * Initiating discussions with accredited certification organizations09:46
willbarnard* Next09:46
willbarnard  * Attend functional safety seminar on Friday09:46
flatmush## Discussion09:46
flatmushAny points to raise?09:46
flatmushI don't know what this underscore 'o' nonsense is, but I'll assume that means no09:47
flatmushSTANDUP ENDS!09:47
paulwaters_ta flatmush :)09:47
noisecellthanks flatmush09:47
Kinnisonflatmush: On the assumption that you're not being disingenuous -- o/ "raised hand" _o_ "hands down"09:48
Kinnisonflatmush: also relevant: \o/ "yay" and /o\ "oh noes!"09:48
persia\o is also sometimes used for "raised hand"09:55
flatmushmaybe this belongs in the wiki?10:00
paulwaters_flatmush, the standup log - yes please10:03
flatmushDo we need to? It's already logged10:03
paulwaters_ah ok - that should be fine then I guess10:09
*** willbarnard has quit IRC10:26
*** willbarnard has joined #trustable10:26
persiaThinking about the block of meeting earlier, and the discussions had about trustable before, has there been clear documentation of t.intent with appropriate attribution and evidence of approvals?11:50
persiaHas that been analysed to develop t.requirements?11:50
persiaIf not, I suspect either a) all the work described as being done above will need to be redone or b) the result won't match the emerging definition of "trustable".11:51
persiapaulwaters: That might be most actively interesting to you, as the person who appears to have brought the meeting to this channel.11:51
paulsherwoodpaulwaters_: need to update the contributors list if others are participating12:09
paulsherwoodpaulwaters_: and pls do take note of persia's comments... for various reasons this seems to be the only place he can land them12:09
* paulsherwood would prefer that the standup notes go to the trustable distros wiki also please12:11 has been reliable for some years, but it's not backed up aiui and i'd hate to see the content disappear12:12
paulwaters_paulsherwood, standup logs will be available here
paulsherwoodperfect tvm12:27
* paulsherwood reads and wonders whether it implies that 500ms guarantees would be enough for a whole range of usecases12:30
noisecellpaulsherwood, wrong link?12:34
persiapaulsherwood: It does not.  The common test is to do something like run a 15fps film on a 30fps display: most folk will see the stutter.  If you run the same film on a 15fps display, lots of people don't see the stutter.  The visual system inappropriately magnifies discrepancies with assumptions.12:36
flatmushdepending on what you're looking at, the human eye can see in excess of 300fps12:38
flatmushI have a 144Hz monitor, you'd be surprised at what it allows you to see that a 60Hz monitor cannot12:38
flatmushbut usually only in a very small area12:38
persiaPut another way, Saccadic Masking only masks the timing errors that are caused by the perceptual apparatus, not any timing errors that occur in other contexts.12:40
paulsherwoodwillbarnard: did you get anywhere with the top-level mermaid diagram?13:24
paulsherwoodnoisecell: no, that was the link i intended13:26
paulsherwoodi was being a bit flippant, though13:27
noisecellpaulsherwood, I read it carefully after persia and flatmush have continued talking about it, sorry about that13:27
willbarnardpaulsherwood: I produced something very basic, edmund was looking to try and tweak it13:27
paulsherwoodwhere is it?13:27
willbarnardit is not published13:28
paulwaters_persia, going back to your earlier points on t.intents - these are covered here ->
*** willbarnard has quit IRC14:20
*** willbarnard has joined #trustable14:30
persiapaulwaters_: Thanks: that finally loaded for me (I have insufficient bandwidth at the moment).  Makes me wonder if we ever formalised the definition of "intent", but covers my first point.14:33
persiaI still think it premature to inspect particular technologies before building requirements, but in practice, the technologies available today may inform the requirements.14:33
persia(and I have yet another reason to try to get a github account when I next have functional bandwidth: the contributor list doesn't match any information I could find elsewhere in the repository, and there is no indication of how "contributor" is counted).14:34
paulwaters_persia, I'm still getting up to speed with the work done to date on the "t." set of definitions, I have a feeling that they may need to be extended for the trustable distros work though14:35
persialast week willbarnard and I were carefully arguing through them: there's some still to be argued, but we started from the premise there were probably too many.14:37
persiaWhat sort of thing do you think would be added?  Can we alternately usefully reuse something that exists if we change the definition a bit?14:38
paulwaters_persia - I think I need to see the diagram paulsherwood mentions above first so I can understand the relationships between the various "t." elements - once I understand this more I'll be able to comment netter15:05
*** paulwaters_ has quit IRC15:06
*** noisecell has quit IRC15:57
*** paulwaters_ has joined #trustable16:02
*** willbarnard has quit IRC16:16
*** paulwaters_ has quit IRC16:19
*** weyfonk has quit IRC17:06
*** toscalix has joined #trustable19:11
*** toscalix has quit IRC23:41

Generated by 2.15.3 by Marius Gedminas - find it at!