(or it's showing you're muted)
<scribe> scribe: manu
<scribe> scribenick: manu
<stonematt> https://lists.w3.org/Archives/Public/public-vc-wg/2019Jan/0017.html
stonematt: We have a long agenda
today, need to be tight w/ timing.
... Here's our agenda today:
2-Unassigned issues [1] (5 min)
3-F2F planning/decision [2] (10 min)
4-Publish Working draft update [3] (5 min)
5-Process to get to CR [4] (10 min)
6-Explainer intro [5] (5 min)
7-PR review (CR Blocker Checkin) [6] (10 min)
8-Test suite checkin [7] (5 min)
stonematt: Hi, my name is Matt Stone - as of last week, I joined Brightlink, a credential management company, credentialling platform and observational testing product. Brightlink is based in Atlanta, GA and is in the process of joining w3.
<stonematt> https://github.com/w3c/vc-data-model/issues?utf8=✓&q=is%3Aissue+is%3Aopen+no%3Aassignee
<cwebber2> not a blocker imo
<cwebber2> oh well manu disagrees :)
<inserted> scribenick: dlongley
manu: Anything having to do with
the test suite I can chime in. #368 is a CR blocker, #367 is a
CR blocker.
... #364 I have to take a look at.
... We could go into CR with them as mistakes if we're
definitely going to do another CR but we should get them fixed
up.
<manu> stonematt: Let's assign them then.
manu: I'll take 364, 367, 368.
<kaz> https://github.com/w3c/vc-data-model/issues/364
<kaz> https://github.com/w3c/vc-data-model/issues/367
<kaz> https://github.com/w3c/vc-data-model/issues/368
<inserted> scribenick: manu
stonematt: Maybe David Chadwick can take this one?
<stonematt> https://github.com/w3c/vc-data-model/issues/363
<stonematt> https://github.com/w3c/vc-data-model/issues/362
stonematt: I'll do 360
<kaz> https://github.com/w3c/vc-data-model/issues/360
stonematt: Is that a review blocker?
manu: Nope, we may need some non-normative text, but that's it.
burn: I agree, input doesn't seem to cover stuff we haven't already discussed.
<Zakim> manu, you wanted to provide status update.
<inserted> scribenick: dlongley
manu: Things are good, we have a venue, it's free and has conference facilities and seating and is in a hip space.
<burn> Propose we continue as decided last week
manu: We have everything we need
except maybe the chairs, we're working that out there. We don't
have food yet but that's a detail that's easy.
... As long as we get that last issue resolved we'll be
good.
<inserted> scribenick: manu
stonematt: What about Chair availability?
burn: I think we should continue
as decided last week - let's plan on this - chances are good
that we can do a passport renewed and travel will be approved
(so Chairs will most likely be available)
... We will create a logistics page to the group
<JoeAndrieu> http://rwot8.eventbrite.com
<burn> VCWG meeting is March 4-5
burn: The VCWG Meeting is March 4th-5th
manu: The meeting is in
Barcelona
... We'll get a page setup for folks later this week.
stonematt: We decided last week to publish today.
<kaz> VC data model draft
<inserted> scribenick: dlongley
manu: The static document was created last week, I link checked it, but didn't pubrules check it because it's supposed to do that ... you should have an email about it. Everything but pubrules is done.
kaz: If you want I can take over and do the pubrules checker.
manu: Thanks, yes please, I'll also check again to make sure there are no issues.
stonematt: Would we expect that completed today?
manu: Today. At least the pubrules part of it.
stonematt: Ok, pubrules complete today.
burn: Can we get it published today?
kaz: It's already 1am in Japan so not sure if I can complete everything. If there is any problem I can talk with Manu again so it might be tomorrow.
stonematt: What I'm hearing then
is that we'll expect the final doc by end of day
tomorrow?
... I think it's ok for you to sleep for a few hours.
... Please send a note to the group when it's published,
Kaz.
kaz: I think I sent a message to
Manu, Dan and Matt yesterday -- BTW, Manu's message went to my spam box and I've just found it.
... I'll send an update to that email thread.
<inserted> scribenick: manu
stonematt: We'd like to cover
process stuff next - we're putting together stuff that has us
getting to CR. Finishing content of Data Model spec plus
everything else that W3C Requires.
... We've shared this w/ a few folks, want to get into general
circulation for WG, keep track of where we are.
... The target (at the bottom) is to shoot for late Jan / early
Feb to vote for transition.
... We have a lot to do between now and then.
... The items in column B are the things we think we've done...
working from top down...
... TAG review need to do that - need explainer done - have a
few CR blockers that we need to address
... We've made great progress, but there are still a few out
there, need to know where we are on those.
... We do need a formal wide review
... We might make the case that we've done decent wide review -
that will happen once we have a spec that is ready for CR
... Then we need to respond to those comments. Column A is a
priority ordered list - big step between responding to external
comments, and identifying at risk features, item 93
... External comments after wide review - as we pull trigger
going into CR... we may have more work between 11 and 93 -
we've gotten a lot of comments from community, so responses
will most likely not be substantial.
stone: Dan, if you want to add anything, please do.
burn: We're going to have to pull
back our target dates... TAG Review is going to take a couple
of weeks. We have to get the explainer done.
... We can say we have had wide review, we've already sent it
to everyone. Those dates, we're not going to make, but we're
not too far off from them.
... Test Suite finalized is still relevant - we do need it
finalized.
... That's all.
stonematt: Does TAG know that this is coming? Anything we can do to give them a heads-up? Will they be responsive?
burn: They probably don't know this is coming, Chairs haven't told them, so advance email would be a good idea.
<stonematt> ACTION: chairs to give TAG a heads up that we will ask for a review shortly
<inserted> scribenick: dlongley
manu: I request that we do that ASAP because it takes them weeks to assign someone. It doesn't hurt to let them know that in the next 2 weeks-month you'll get a review request ... please be ready. It's pretty much ready for TAG review now. From what I can see so far, there's no substantial changes that will happen from a technical architecture standpoint between now and CR.
burn: That's correct it's just that we're waiting on the explainer ... which is a gating function to get anything done by TAG now.
<inserted> scribenick: manu
oliver_terbu: What is needed from the JWT end? Beyond the two different implementations?
<inserted> scribenick: dlongley
manu: So the feature at-risk thing is just something we'll write in case we don't get two independent implementations. Any feature like that where that's a concern should be marked at-risk. We do need two different interop implementations for everything but we need test suites. So we'll need that for JWT stuff there.
<manu> https://github.com/w3c/vc-test-suite/blob/gh-pages/test/vc-data-model-1.0/50-jwt.js
manu: I refactored some of the
test suite yesterday or the day before -- there is now... the
file in IRC.
... Any MUST statement in the JWT section of the spec has to be
tested. You have to put in a test that tests each JWT feature,
the conversion mechanisms and that kind of stuff.
... I expect 10-30 tests that have to be written there to
demonstrate JWT VC interop.
oliver_terbu: What is the deadline?
manu: I think we were trying to
get the entire test suite done before entering CR so we knew
what we were testing. But this is up to the chair ... one
answer is test suite is done before going into CR... that would
mean deadline is next week or two.
... The other way to do that is "we're doing testing in CR so
we don't need it for a month".
... But having it going into CR is usually better.
<inserted> scribenick: manu
burn: From a Chair perspective,
we want to see those tests in before CR.
... It's just a lot cleaner to describe the process.
... wrt. the feature at risk - if you don't mark something as
feature at risk, pulling it out of the specification requires
another CR cycle.
... marking something a feature at risk makes it so you can
pull it out of the spec w/o having to go through CR
again.
... We should make every effort to get what we can in before
CR.
bigbluehat: I've been working w/
Manu on test suite an did file two issues that are CR blockers.
I want to make sure I'm careful, I'm going to find more issues,
I don't want that to hamper CR.
... Can we, process-wise, address them through something other
than CR-blocker.
<Zakim> manu, you wanted to respond.
<inserted> scribenick: dlongley
manu: First -- big deep thank you
to Benjamin for working on the test suite. This has moved us
about 2 weeks forward in our schedule because of his efforts,
so thank you. We're going to have to go back and forth.
... really quickly so everyone knows what's going on. A lot of
this is fine tuning -- we have to make sure things are testable
and some of the MUST statements may be difficult to test.
Benjamin is doing the group's homework and making sure things
are testable. I don't know how to do this other than you raise
issues and I work with you to resolve them before the next
meeting so we can avoid CR blockers.
... I've already made a pass through the entire specification
so that all of the conformance statements are about the
document format not about how issuer/verifiers respond.
... There were some I didn't know how to test, but I'm hoping
there aren't too many of those.
... As long as the group trusts Benjamin and I to work it out
we'll get through it and knock out what we can in the next week
or two.
stonematt: Would it make sense to add a label for test suite issues? They may/may not be CR blockers but a way to quickly identify on our issue list that this isn't work in the data model necessarily as a way to differentiate that?
manu: I'd rather not add more process at this point, I think we can work through it more quickly, adding more process will slow it down.
stonematt: Ok.
... Dan, something about document conformance?
<inserted> scribenick: manu
burn: I agree that we should try
to not add more process... but if we fall into trouble, we can
make things as "test suite" changes.
... We want initial feedback, even if tests aren't perfect.
Let's try a light touch first...
... I agree w/ Benjamin that we don't want this to turn into a
two month test process.
<Zakim> burn, you wanted to add (perhaps)
burn: The other comment that Matt
is referring to, the official statement about conformance still
talks about "old conformance text".
... we need to get rid of statements about processor vs.
document
manu: +1 to that
stonematt: We need an official
list of implementers that are planning on doing an
implementation.
... I'll put together a document for that, but we think there
will be uPort, Learning Machine, Sovrin, and Digital
Bazaar
... Anything else on process to get to CR?
<stonematt> https://github.com/w3c/vc-data-model/blob/gh-pages/VCDMExplainer.md
stonematt: There are some items
missing related to JWT, and other details wrt. spec that TAG
can look for.
... That is our highest priority now... TAG is a big rock in
the stream to CR.
... What do we need to do to get that done?
<Zakim> stonematt, you wanted to say need list of implementors
tzviya: I put together a list of
what is missing.
... I couldn't find a link to the discussion of JSON vs.
JSON-LD. I need someone to write a short explanation of ZKP. It
would be good to add a list for features at risk.
... Let's see if we can do a section on implementation and
reviews - this is about 10 minutes of work here. We need folks
to step up to do that.
<burn> Don't need JSON vs. JSON-LD issue pointer, just explanation that we have both formats
tzviya: I can volunteer people.
oliver_terbu: I can provide input to JWT section... JOSE vs. JWT - not that, goal is just to provide input.
burn: The goal is to fix it
... We don't need a lot of text, we just need key points
mentioned. JWT support and ZKP support.
<stonematt> https://github.com/w3c/vc-data-model/pulls
<Zakim> manu, you wanted to step through it.
<inserted> scribenick: dlongley
manu: Everything that is left is
editorial from Grant and we're waiting for Ganesh to come back
from vacation to process 313.
... I don't think the group needs to discuss anything, we
pulled in Oliver's PRs on JWTs.
... We should talk about the CR blockers. The only thing we're
slightly concerned about is this last minute review we need to
do for the ZKP stuff. I think Dave Longley started looking at
it this morning and I still need to look at it and make sure
everything is aligned on the ZKP front. Think we're good on
JWT. Everything else impacts the test suite.
... Everything else is minor comments we want to address.
There's an example context URL, what are we doing with that,
that sort of thing. There's a clear answer just have to sit
down and deal with it. From a normative change perspective I
don't expect anything Earth-shattering/ground breaking coming
up. Just need to button things up.
<stonematt> https://github.com/w3c/vc-data-model/issues?q=is%3Aissue+is%3Aopen+label%3Areview-blocker
stonematt: When we started this
process of labeling issues as CR blockers we had another tag
for review blockers.
... There are a couple of review blocker issues still there --
I think these have been disposed of, can we verify that these
two review blockers are done and get these closed?
manu: I will look at those ...
looking at review blockers wasn't part of my editorial
cycle.
... I don't expect any issues to come up with ZKP -- the other
one is about terms of use, I think that's addressed but I will
check to make sure.
... To be super clear, I think we can ask for review, I don't
expect major things to happen to the spec, so people reading it
can read and understand what's going on as long as they are
aware of the issues. They should be aware of open issues so
they don't submit ones that already exist but by and large spec
is ready for reviews.
Brent: I just want to agree, I think we can get rid of the review blocker on that issue but I don't think the issue itself is done because we need to make sure we have what we need in the test suite but the tag can go.
stonematt: Gone!
<Zakim> bigbluehat, you wanted to say P
<bigbluehat> https://github.com/w3c/vc-test-suite/pull/6
<inserted> scribenick: manu
bigbluehat: I'm working my way through the rest of the document now, positive/negative tests - will try to get that fleshed out later this week.
<burn> This points out why test writing should be done before CR - it points out spec wording issues
bigbluehat: Going to try to get things fleshed out in test suite - will help w/ coding in second pass, will try to draw lines on page.
<stonematt> +1 burn
<inserted> scribenick: dlongley
manu: Thank you very much,
Benjamin. When Benjamin volunteered to work on the test suite
it wasn't critical path but now it is -- you and I will have to
go back and forth on this to make sure we're not stepping on
each other's toes, let's make sure we sync up.
... Thank you again for taking that action item. And, if
there's anyone else that wants to volunteers to write tests
they aren't hard -- if you can write JavaScript you can write
the tests.
<oliver_terbu> i will provide tests for jwt
bigbluehat: If you've ever done JavaScript testing it's just mocha and chai if you've ever done that sort of stuff, it's pretty obvious.
<inserted> scribenick: manu
stonematt: That's it for today, thanks all!
<stonematt> thanks all!!