W3C

– DRAFT –
WoT Profile

26 January 2022

Attendees

Present
Ben_Francis, Cristiano_Aguzzi, Kaz_Ashimura, Michael_Lagally, Michael_McCool, Ryuichi_Matsukura, Sebastian_Kaebisch, Tomoaki_Mizushima
Regrets
-
Chair
Lagally
Scribe
kaz, McCool

Meeting minutes

Agenda

Lagally: want to talk about roadmap and timeline
… and then continue the requirements discussion
… incl definition of out-of-the-box interoperability
… would like to use issue tracker to resolve requirements point by point rather than opening general discussion

McCool: would like also give a status update on self-description - plan to resolve

minutes

Jan-19

Lagally: approval of minutes of previous call

McCool: changed the irc, so under wot-profile now

Lagally: typo fix; timeline and schedule; link to requirements on github (clean up of *'s)

Lagally: can we approve the minutes?

Lagally: hearing no objections... approved

roadmap

Lagally: pull out into powerpoint to edit...

<kaz> 1. clarify requirements

<kaz> 2. profile scope, identify gaps

<kaz> 3. contributions to resolve normative gaps

<kaz> 4. contributions to resolve informative gaps

<kaz> 5. publication of second working draft

<kaz> 6. plugfest/testfest

<kaz> 7. incorporate review feedback from 2nd WD review

<kaz> 8. publication of candidate recommendation

McCool: the schedule really can't be extended further; also, I suggest we make the requirements as narrow as possible
… KISS principle; the simpler the spec is, the easier it will be to get it done (and through wide review....)

Lagally: edits points on roadmap to add suggest durations

McCool: proposed schedule puts finalization of normative gaps around time of first plugfest
… so we could at least do some preliminary testing in that plugfest

Lagally: so end of March we should have a sync point
… then clean up some informative gaps
… and a second working draft around mid-May

McCool: note second test fest should focus on testing and generating an implementation report, which implies we need stable implementations a few weeks prior

McCool: need to keep things narrow to hit this

given our current velocity is less than one pr per call
… so we would need to have 23 calls to empty the current prs

Lagally: well, some of these are obsolete and will go away
… we are doing a second loop

Lagally: but I hear you, you feel this is not realistic

only realistic if we review and close PRs out of meetings

McCool: suggest compromise where we use labels to identify easy cases, i.e. things that are obsolete, things that are not contentious (editorial)

Kaz: should be recall deadline of charter period itself
… and consider how long things should take; can go bottom up
… for example, if need to finalize normalize features, and only have 1mo, then we need to shorten requirements

McCool: note that for first plugfest I was only talking about initial testing, not full testing, which will not be feasible

requirements

Lagally: had a presentation Dec 16

<mlagally_> wot-usecases PR 156 - Insert digital-twin-use-case.md and use-case-opc-ua-binding.md into index.html

<mlagally_> https://github.com/w3c/wot-profile/blob/main/contributions/2021-12-16-WoT-Profiles-discussion%2Brequirements%2BOOTBI.pdf

Lagally: see also issue #156
… profile requirements

McCool: still think o-of-the-box-interop is primary requirement, the rest are supporting

Lagally: let's defer that discussion

Lagally: disagreement about some requirements, e.g. human readability

McCool: I also disagree that this should be a requirement

Lagally: let's then create an issue to capture input on that offline

Sebastian: also a minus... have discussed this in the past

Lagally: now developer guidance

comments on the issue, cristiano and I both said we supported it

McCool: for "multiple profiles", is a bit confused; one TD supports more than one?

no, "composable" profiles is next point

McCool: my main concern with both multiple and composable are future extensions; for now we should focus on just one

Sebastian: current mechanism for TD already allows for selection from a set of profiles

McCool: so is a requirement, but already satisfied

McCool: also composable only makes sense if there are multiple profiles

left some feedback on reasons for when we should have multiple profiles
… but this is about the *mechanism* to support more than one, and I agree we should have that

Lagally: regarding composable profiles: let's create an issue to discuss

Lagally: next, verifiable TDs

regarding filing issues to discuss, let's make sure we clean up old issues on these

Cristiano: support validatible requirement

Lagally: yes, on identification; is moot if we can't identify them
… 4m left, will commit on a branch and make a PR for discussion

McCool: would be nice to link issues from requirement docs so it is easy to find

Lagally: ...

McCool: let's not merge just yet but keep as a PR so we can capture comments

ok: PR #179

Lagally: PR #179, leaving open

Lagally: also, there is a label, "requirement clarification" for requirements discussions
… and please, let's close on this as soon as possible

<kaz> (after the discussion, Kaz asks about how to proceed, and Lagally explains his plan based on the "Roadmap")

Lagally: planning horizon at this point just first three points on plan

<kaz> [adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).