* i dont think so... we could switch to skype or hangouts?
* Nick is on a plane i think and dont think we can resolve this easily.
* i was able to dial in the the australian number so the meeting itself is valid
* ok coming up
* try this https://hangouts.google.com/call/MbTSvGuKXd1X73Rv4C81AEEM?no_rd
* you show up but i cant hear anything
* OK
(it's pretty full, let's see how far we get)
https://www.w3.org/2017/dxwg/wiki/Meetings:CNEG-Telecon2019.09.12
+1 (I wrote it...)
+1
Resolved: Agenda confirmed
https://www.w3.org/2019/08/29-dxwgcneg-minutes
+1
+1
Resolved: Minutes from last meeting approved
https://www.w3.org/2017/dxwg/track/products/4
Action-193?
<trackbot> Action-193: Rob Atkinson to Move jmeter test suite to within w3c systems -- due 2018-09-05 -- OPEN
Can you say anything about that, roba?
both still open - though 358 probably moot now - I'll coordinate with Nick during implementation when I have choices to make - we can provide any changes needed back as comments.
OK
See steps to recommendation https://github.com/w3c/dxwg/wiki/Steps-to-Recommendation-2019
the last PR frees me up to complete the HTTP part of the implementation I'm doing
which is the last PR, roba?
https://github.com/w3c/dxwg/pull/1066
OK, fine.
Ready to move on to the next topic?
ok
See steps to recommendation https://github.com/w3c/dxwg/wiki/Steps-to-Recommendation-2019
We need to ask the group for review and possibly other communities, too
… Do you know if Nick is available tomorrow?
Yep - OGC can be wider than the CSW spec WG - I'm about to present on the benefits of it to a session at OGC in a couple of hours, and I have a contact at the US NGA interested I will urge to look at it.
… possibly - he arrives back in the afternoon apparently
Can you add those to the list in today's agenda?
Obviously we need a final editorial clean up to flag "features at risk"
ok - edit agenda?
Which features would you see as "at risk"?
IETF and the registration of the headers itself I guess as we dont fully control that..
That's an interesting point. Do we need to show implementation evidence
… or do we need to show that they have been standardised?
This community has little idea what they mean by profiles but conneg makes it clear what the implications are so that should be an issue - if they cant define conformance maybe they cant use this spec, but that doesnt invalidate the spec
we do - but not till end of Oct I think
When you say "this community" do you mean "IETF"?
no, DXWG
OK, so let's mark the http stuff as "feature at risk"
… I can take an action on that
I cant see how to avoid it..
Action: LarsG to mark http implementation as "feature-at-risk" since we don't know what IETF will say about it
<trackbot> Created ACTION-371 - Mark http implementation as "feature-at-risk" since we don't know what ietf will say about it [on Lars G. Svensson - due 2019-09-19].
If we implement the rest, and we have no counter-proposals active for an alternative (or proof it wont work) then I say its not at-risk
… as there are no external dependencies. Functionally its a very light spec.
OK, so are you confident there will be two indenpendent implementations of QSA?
Yep Nicks Australian Government Linked Data stuff and my OGC stuff - using different code bases and strategies.
Great
Regarding review, I can offer to contact IETF, DCMI, Europeana, DCAT-AP.
Shall we ping Nick that he contact OAI-PMH, ODRL and tool makers?
<silence>
there was a comment on issue https://github.com/w3c/dxwg/issues/1055 where people are saying that DCAT-AP doesnt support content-negotiation for data distributions (in general) - claiming DCAT specifies this
Yes, but is that an issue for our spec?
but its incorrect as it clearly states it supports access via API and the cardinality of dc:format is not restricted.
So no - i think its an issue for DCAT-AP - but if they really do hold that position then this spec isnt relevant to them.
(it makes no difference to this spec)
OK, but we can ask them through Andrea anyway. I think he knows enough of the context
I suspect its just misinterpretation and some slightly inconsistent wording in DCAT.
+1
I'd like to move on to review of the public comments list. OK?
Andrea will care about other APIs like OGC Web Feature Server which serve multiple formats..
OK
https://lists.w3.org/Archives/Public/public-dxwg-comments/
Are there any open comments we need to address?
There is https://lists.w3.org/Archives/Public/public-dxwg-comments/2019Aug/0004.html from TomB
… this is repeated in #1017 where Nick has asked (https://github.com/w3c/dxwg/issues/1017#issuecomment-526164813) TomB and Karen if they are fine with the response
… no comments so far
… so no, I haven't got a response from TomB
… I'll ping them on the comments list
did you get a response from Tom after answering all his questions?
Action: LarsG to ping Karen and TomB regarding the answer listed in https://github.com/w3c/dxwg/issues/1017#issuecomment-526164813
<trackbot> Created ACTION-372 - Ping karen and tomb regarding the answer listed in https://github.com/w3c/dxwg/issues/1017#issuecomment-526164813 [on Lars G. Svensson - due 2019-09-19].
I think that apart from that we have no open comments we need to respond to
At some point these conversations drift off into lack of clarity about what they mean by profiles and the conformance implications - which are not actually relevant to this spec - so how do we draw a line under these threads
That's difficult. We need to show that we have addressed all original questions coming in through the comments list
… Questions and comments coming through GitHub are solved when we're allowed to tag them as "due-for-closing"
Can we move on to open GH issues?
ok
https://github.com/w3c/dxwg/issues?q=is%3Aopen+is%3Aissue+milestone%3A%22Conneg+3PWD%22
7 open...
… in this milestone
… 19 as a whole: https://github.com/w3c/dxwg/issues?utf8=%E2%9C%93&q=is%3Aopen+is%3Aissue+label%3Aprofile-negotiation
Let's start from the bottom
#290 https://github.com/w3c/dxwg/issues/290
well 1031 - PR has now been done it can be closed
i dont have that im my list of 7 ?
There is emerging consensus on #290 (three +1 votes), if you add yours, it's four
(no it's in the list of 19)
done - its a requirement issue anyway - just a check off its covered.
#389 https://github.com/w3c/dxwg/issues/389 what can we say about the OGC implementation?
in progress - been wiating for finalisation of HTTP token mappings which are necessary
OK, then you should be ready to go now!
387 can be closed - I'm ready to go..
#505 https://github.com/w3c/dxwg/issues/505
We haven't reached consensus there yet
… perhaps we cannot give normative advice
i dont think there are any arguments for HTTP to take precedence over QSA are there?
so IMHO we just need to check its clear
No there isn't. In my view the URL trumps any http header.
… can you make a text proposal?
(needs to be merged by Monday...)
OK
Action: roba to provide text for order of precedence (#505)
<trackbot> Created ACTION-373 - Provide text for order of precedence (#505) [on Rob Atkinson - due 2019-09-19].
#546 https://github.com/w3c/dxwg/issues/546 RobS hasn't answered to Nick's comment
We're well past the hour... How long can you hang on?
as long as you wish
this issue is moot as list tokenbs was removed long ago..
OK, then let's continue!
#548 https://github.com/w3c/dxwg/issues/548 Nick needs to look at that one. I'll ping him
#549 https://github.com/w3c/dxwg/issues/549 I can check with RobS on that one
Actually also #548 also solved by PR 1031
Oh good! I'll add that.
#575 https://github.com/w3c/dxwg/issues/575
Here there's still work to do (listed in https://github.com/w3c/dxwg/issues/575#issuecomment-530362099)
- sorry just catching up on 549 - its a "nice to have" we might leave as an open issue pending availability of a canonical representation?
Can you take care of Antoine's issue 15 since it's about 303 redirects?
Yes, we can tag #549 as "profile-negotiation future work"
PROPOSED: Tag GH #549 as "profile-negotiation future work"
+1
+1
Resolved: Tag GH #549 as "profile-negotiation future work"
I'll ask RobS if that's fine with him
#678 https://github.com/w3c/dxwg/issues/678
I think we need to have an action on editorial for #575 - but I dont see any substantial changes are required - just better examples and explanation (we've updated several examples to use http-range14 friendly options anyway)
OK, can you take that action?
I've been hammered by deadlines this week - will try to find some time on weekend
Action: roba to perform final editorial changes from #575 (https://github.com/w3c/dxwg/issues/575#issuecomment-530362099)
<trackbot> Created ACTION-374 - Perform final editorial changes from #575 (https://github.com/w3c/dxwg/issues/575#issuecomment-530362099) [on Rob Atkinson - due 2019-09-19].
Thanks, roba!
OK, back to #678. Kam hasn't responded to Nicks question. We should ping him again. I'll do that
yep nothing to do otherwise
#976 https://github.com/w3c/dxwg/issues/976
I cannot determine of this issue is resolved yet. We should ask antoine about that.
… pinged him in the issue.
Its another rambly discussion that doesnt seem to have any effect on the spec - communities can decide what they call profiles and make up any conformance rules - conneg just provides a means to deliver what they ask for..
#1002 https://github.com/w3c/dxwg/issues/1002
I've pinged antoine if the issue has been discussed to satisfaction
again - its just exploratory questions - not a concrete issue
OK, and it's not tagged for the milestone
#1017 https://github.com/w3c/dxwg/issues/1017 we had that one further above with a corresponding action
#1022 https://github.com/w3c/dxwg/issues/1022
Antoine has re-opened saying that "the current text (especially the hyperlinks) do blur the differentiation between functional and data profiles."
OK antoine has said this a "potential showstopper"
Yes, but that seems to be partly terminology. IIUC it could be fine to call the document "Content Negotiation by Data Profile"
It could be fine - but its actually fairly obvious. "content negotiation" implies data
True, although not all content is data...
… depends on your definition of "data" of course
#1031 https://github.com/w3c/dxwg/issues/1031
There is a description and examples in the document, so perhaps we can mark this as due-for-closing
yep PR was merged
1041 is dealkt with - Nick Raised it so needs a statement from him it addressed.
1064 is an editorial cleanup we can close when PR removes note.
I can take care of #1064 now that #1063 has been merged
+1
which leaves #1042
OK, done with issues
I can do that
Action: LarsG to update changelog in ED
<trackbot> Created ACTION-375 - Update changelog in ed [on Lars G. Svensson - due 2019-09-19].
AOB?
<seemingly not...>
can we discuss #1042 - I think it might address the issue Antoine has by showing how its necessary in practice
i.e resolve 1022 ?
… i.e. how and why data and functional profiles are necessary in the grand scheme of things - as well as provide guidance for how to describe a service offering conneg-by-ap
I'll create a PR and ask Antoine to review ?
Yes please, and I'd like to review, too (then perhaps I understand what's meant...)
of cource you and Nick must review too!
Action: roba to create PR to resolve #1042 by showing how data and functional profiles work together
<trackbot> Created ACTION-376 - Create pr to resolve #1042 by showing how data and functional profiles work together [on Rob Atkinson - due 2019-09-19].
its not a feature, so informative appendix - but perhaps might cut through the abstractness
zakim says we're done!
OK, so it wouldn't be normative content (i.e. it's editorial)?
hmm - open minded on that - its significant example
The point is that it's important for the review process
… if it changes or adds normative content, we must highlight that
yep
(IMHO)
no its example DCAT metadata illustrating use of (and nature of) the idenfied functional profiles, not a requirement
OK.
So shall we call it a day?
(if we cant get the PR through we just dont provide an example)
yep - good work - thanks!
Thank you! Sleep well
Bye
Succeeded: s/is/i.e/
Maybe present: PROPOSED