See also: IRC log
<azaroth> Gah
<ShaneM> trackbot, start meeting
<TimCole_> scribenick: bigbluehat
<TimCole_> PROPOSED RESOLUTION: Minutes of the last WG call are approved: https://www.w3.org/2016/06/10-annotation-minutes.html
TimCole_: any concerns about the minutes?
<azaroth> +1
+1
RESOLUTION: Minutes of the last WG call are approved: https://www.w3.org/2016/06/10-annotation-minutes.html
TimCole_: anyone have a follow-up from the CR call?
ivan: there is a minor issue. azaroth do you want to explain?
<azaroth> I created: https://github.com/w3c/web-annotation/issues/313
<TimCole_> Tag issue: https://github.com/w3ctag/spec-reviews/issues/93
ivan: there is an issue form the TAG which needs addressing with regard to the Protocol
azaroth: Erik and Mark do seem to agree on this one.
TimCole_: most of these we did look at with
Erik's comment
... and we thought it had been resolved
TimCole_: so it resurfacing was a bit confusing
ShaneM: it looks like they kept kicking the
can down the road and it also fell of their radar
... I think we can close the loop with them
ivan: azaroth will you contact them? or what
should we do?
... I would really like to have this solved
... the fallback is workable...but it is a big hit
TimCole_: I was trying to find the docs from
our chat with Erik
... but I can find it
azaroth: Erik is `dret`
TimCole_: "avoid constraining HTTP"
<TimCole_> https://github.com/w3c/web-annotation/issues/51
TimCole_: I don't know if Mark looked at that and still had a concern?
azaroth: we closed issue 51 on September 9th
last year
... Mark's comment was in November
... Mark may want more changes along those lines...but it's hard to tell
TimCole_: he may not have seen the updates
after we made changes that followed handling Erik's comments
... it may be if we show him the latest draft, he'll be fine with it
azaroth: ivan from a process perspective
... if we address these issues
... it would be a normative change if we remove normative
requirements...
... would we need another call for consensus?
<ShaneM> I recommend we don't borrow trouble.
ivan: you mean in the working group
... we should be able to do that via email and not bind it to a telco
... if there are changes you agree with Mark
... and we put it in an email and give it 5 or so days
... and if no answer, then we accept it
... that should be enough
TimCole_: anything else to say about the CR call?
ivan: no. I have all the documents ready
... there is one we'll have to change slightly, but that's a minor thing
... I'll send out a request to the web master to check them
... some of the issues were taken care of by azaroth
... so I think we're done in this respect
... of course if there's a change in the protocol, then it has to be
checked again
azaroth: timing wise, I'm on vacation for
most of the next couple weeks
... and I am forbidden to be connected from the beach
ivan: if this goes wrong, it could push the protocol publication out to august...
TimCole_: let's be optimistic
<azaroth> (specifically June 29-July 4)
ivan: perhaps you can contact Mark privately
first
... we thought this was closed...so some sort of a leeway would be
helpful
... try to find out what the minimal amount of change is
... and if you can do that today, that would be great
azaroth: right. I'll do it right after the call
ivan: ideally and answer by close of business Monday would be helpful
TimCole_: there was a question about call
timing
... should we meet weekly or adjust the frequency
... this relates to comments from a few people
... who I think are interested in implementing the protocol
... they're not currently members of the WG
... but we want to encourage them to implement
... First question is, should we reduce frequency of the calls during
July and August
ShaneM: I don't think we should until we're done with testing
TimCole_: that was my thought also that we should dedicate our calls completely to testing
<azaroth> +1 to staying weekly until we're no longer dealing with issues
TimCole_: what is the policy to invite folks to calls no in the WG?
ivan: it's totally fine. they don't have a voting right, but if they have input, there's no issue
TimCole_: we may just skip next week then and meet again on the 8th
azaroth: I would prefer to keep it at full
working group every week
... and if people don't show up, then we cancel
... this would be better than explicitly going to ever other week
... it's easier to cancel
TimCole_: k. let's stay with every week
... I'll try and get a sense of who's on the calls
... we'll keep the focus on testing and implementing
... that work?
azaroth: +1
TimCole_: do you want to talk about the issues? or go right to testing?
azaroth: if we could chat about the issues, that'd be great
<TimCole_> https://github.com/w3c/web-annotation/issues/308
azaroth: these issues are both by the same person
<TimCole_> https://github.com/w3c/web-annotation/issues/309
azaroth: are there things we could do to not
shut down the discussion, but be clearer that they've already been
discussed
... and acted on
... they are both marked as `postpone`
ivan: right. that's what he's contesting
TimCole_: k. then lets discuss that
azaroth: 308...
... should we ask for specific improvements?
ivan: I don't completely understand what he's wanting
azaroth: I haven't quite sorted it either
TimCole_: it looks like things we've already
decided upon
... the type vs. format thing has been discussed before
... they confuse people who are not familiar with Dublin Core
... this has been a confusion for greater than 15 years...
... I don't think we can fix it with an editorial change
azaroth: I don't think there's a solution that will make him happy
TimCole_: anyone else have thoughts?
PaoloCiccarese: can you explain the last
line you said?
... it seems its mostly a change in terminology
ivan: right. all of them are
PaoloCiccarese: he claims at one point that
information is redundant
... azaroth you said ExternalResource is wrong?
azaroth: yeah...I don't believe his example
is correct
... because you could do text inline as a TextualBody...and may not be
external
... so I believe its correct as it stands
PaoloCiccarese: I think he's making a
distinction between embedded and external resource handling
... but I think that's the distinction he wants
... I don't think that's really crucial
TimCole_: this has been discussed elsewhere
... I think we should go ahead with it as is
... we're doing what people have been doing with this issue for years
... do we need a vote?
azaroth: TimCole_ can you leave a comment
TimCole_: sure
... so. 309
<azaroth> https://github.com/w3c/web-annotation/issues/309#issuecomment-227387975
azaroth: one request: can we add example
extensions to the body
... we could make the prose a little clearer, perhaps, but you could do
these extensions anywhere
... so my suggestion is not
... and with the other, there doesn't seem to be a need that isn't
addressed by what we've already discussed
ivan: agreed.
PaoloCiccarese: I'm reading his comments now
... it seems he's wanting TargetResource to be renamed to
ExternalResource so it's consistent with their meeting
... but I certainly agree with azaroth that if you want to extend
something, then just extend it
... we don't need to provide examples
... and can't for all the cases
azaroth: yeah. ExternalResource is really
just Resource
... and you'd talk about it the same way you would any other resource
TimCole_: for right now, I don't think there's anything we could change to make this clearer
azaroth: getting another example from them
might be helpful, but I don't think there is one that would make this
compelling enough to change
... and considering the protocol in the mix, it supports Turtle which
doesn't have named graphs
ivan: if named graphs is the need here, then
we could require support for named graphs via TriG--which is standard
now as of the latest RDF spec
... and we could require that
... but we're at feature freeze
PaoloCiccarese: with regard to named graphs,
there's really no way around wanting these
... you need them.
... but I completely understand the feature freeze
... and the idea of moving from Turtle to Trig would be nice
... framing could be another one.
ivan: named graphs in JSON-LD are doable,
but very ugly
... and if we want to push them back into the spec and keep it
accessible to web developers, then it'd take far too long
... so version 2. Postpone.
azaroth: given that it's JSON-LD, they can
use the `@graph` stuff and extend it however they'd like
... it just won't be interoperable
ivan: I would advise you to either rename
the issue
... so it's clear that we'd need to incorporate named-graphs into the
specs
... and it's clear what we're postponing and discussing
... I'm noticing bigbluehat is asking for changes on the protocol
document
azaroth: there's an example that's wrong or
the text is wrong
... in response to creating a resource
... right now the examples in response to creating an annotation refer
to the container
... Link headers refer to the request IRI
bigbluehat: I think the example is correct, but not the prose
... so the prose needs to be updated.
... these errors are coming to the surface as I do testing
ivan: so the protocol is in a slight delay, but can you please handle today, so that what goest to Mark and Webmaster is final
... please generate final doc in both main directory and admin
azaroth: Benjamin and I will resolve after the call
ivan: wants to make sure what he puts up is final.
ShaneM: sadly, I don't think I'll be around
for the next call
... I send in my availability
ivan: wasn't there some sort of thing you needed from us?
ShaneM: yes. in order to get this into WPT
it needs peer reviewed
... we need others besides bigbluehat to take a look
... so its clear that the WG agrees
... with this approach
TimCole_: for time frame?
ShaneM: two weeks ago.
... in that we can do any actual testing until it's done
TimCole_: we are still working on the tests
ShaneM: right, but we need the testing
framework up and running
... and all the tests have to go through review also
<ShaneM> https://github.com/w3c/web-platform-tests/pull/3173
<ShaneM> this is the PR that we need peers to comment upon
TimCole_: I'm still working through the $ref bits in ajv
ShaneM: thanks for the email this morning. I
think it's fine to go ahead...
... actually I've been doing meeting since I saw it
... it looks like you want to pull in pre-defined
... which is fine
... I'll add that to the code and update the pull request
TimCole_: at that point, I'm fine with it
... there were some questions about JSON Schema version 5
<azaroth> +1 to using v4
ShaneM: I'd thought I'd handled this in email, but I thought I'd recommended sticking with v4
azaroth: I'm looking at mixing this into my
python implementation
... and the JSON Schema validator I'm using only supports v4
TimCole_: it's not been clear until now that
the PR is ready for review
... but it seems like it is, so we'll work harder to get you the review
... assuming you get comments back from the WG
... when do you think we can get this merged?
ShaneM: well, it's ready to be reviewed
... but I've never done this before
... so I'm not sure what the process is or how quickly the owners of the
code will act
TimCole_: I'm sorry for the lack of the
response from the WG
... but if you can ask again, hopefully you'll hear back
<TimCole_> scribenick: TimCole
<TimCole_> bigbluehat: finishing out basic protocol testing
<TimCole_> ... server will give you a container and let you run tests
<TimCole_> ... container will time out after suitable interval
<TimCole_> ... will come with some html to get you started
<TimCole_> ... once it's in place will need peer review