billroberts: Any comments on the minutes?
Resolved: Previous minutes accepted
billroberts: Catching up with the threads
billroberts: That's the latest editor's draft
… Jon has done some useful work on this in the last 2-3 weeks.
… Some things could do with a tidy up, like references and boiler plate text
… But in high level terms. We were going to propose this as a FPWD, but had a discussion on where the actual spec should go. Decision was to keep spec separate
… May then go through proper standardisation process. Now just referenced from this doc
billroberts: Still to do - references
<kerry> phila: a custom sentence that give an indication of maturity -- comments sought, how many more iterations..
phila: Explains the SotD section
Action: phila To investigate what's required for privacy etc, for non-Rec Track docs
<trackbot> Created ACTION-289 - Investigate what's required for privacy etc, for non-rec track docs [on Phil Archer - due 2017-03-22].
billroberts: Apart from that, the doc is near to complete so I'm happy to put it forward as a FPWD
kerry: Just looking at it... section 6 looks like an appendix. The table is good - just the section looks like we wrote it because we had to
… Something about satisfaction of UCR, reference to UCR etc. Mapping sounds too process-driven
kerry: Actually, it's more the word document I don't like. I'd like it to be more positive
roba: It's the word document that jumps out to me
kerry: We have developed UCRs for coverages
roba: Either call it use cases and Reqs or just Reqs
billroberts: I wanted to mention UCs as wew have an output called that, but happy to remove the word document
phila: Actually, that DWWBP table was auto-generated
kerry: Another thing - the BPs... how do we deliver or not to the BPs
… That doc has a fair bit to say on the topic
billroberts: I could add a section to talk about that
roba: Can I suggest that we try and deal with this in the F2F. This wasn't available when the BP doc was written
roba: If we're claiming that this is a BP then the BP doc should back that up.
… We need to reference stuff in the future work section
billroberts: Both docs are evolving, yes
<billroberts> specref: it's great!
kerry: do we need to say/do anything about Maik dropping out of group? Think somewhere in document?
<kerry> billroberts: comments on QB4ST , interested in dggs
billroberts: Have you had time to read the comments I sent
sam_toyer: Thanks for sending the comments - very helpful. Not had a chance to incorporate them yet
billroberts: On the implementation section - strawman of current DBs good idea. Current set up does handle filters in reasonably intelligent eway
… Goes into the DGS
… Quite fast. Can view DDGS as a kind of index.
billroberts: Interesting. I think it woujld be good to add something on that to the document. That provides insight into why using the DDGS provides efficiency
<roba> qb dimensions are usually indexes
billroberts: So I made a pull request of just trivial changes
… my comments worth looking at are in the email. Stuff in the PR are mostly typos and minor wording changes
… Please look through and merge (or I can)
kerry: I can do that
billroberts: Anythign more on EO-QB?
… Anything you want us to talk about in Delft?
kerry: I don't think we'll have a significant increment for the F2F
… The main thing we needed to do is the mapping
… More work on how the client works with it and more on the BPs. I don't think wew'll have that in a good enough state to put up for a vote
billroberts: what's the current end of WG life date?
kerry: 30 June
billroberts: So we have 3 months
… They'll pass quickly for all of us.
… So you're thinking one more iteration?
kerry: Yes. I was hoping this would be it but we're going to miss it.
billroberts: Any more of EO-QB?
billroberts: Just to note - I had a proper look the other day and sent some notes and a pull request which, again, is very trivial and Rob's already responded.
… There's a place holder for an example section - I think that would be very useful
… Rob said he was hoping the EO-QB stuff would provide those examples
billroberts: A lot of the doc is quite abstract but the principles are clear enough
… Examples would help make it less abstract
roba: The actual example I was looking for was more ?? I'd like to link it up more with CoverageJSON
… I was wondering whether there was a more canonical example where we use CoverageJSON
… ANd then I can give an example of what the QB4ST metadata would look like
… I can look again at the DDGS hierarchy
… I can havea a go at doing that in conversation with you
billroberts: Sounds interesting but I think we should give Jon Blower a shout
roba: If you have real world services... it's easier to look at... describing an existing non-RDF thing that has harder slicing and dicing
… where there are APIs and services for the hierarchical data
roba: Once it's in the more flexible RDF version you lose the challenge
billroberts: I'll ask Jon for a simple but reasonably representative example
billroberts: I think what's important is taking the RDF QB principles...
roba: It's a lightweight vocab for doing these things
roba: Let's see if we can get that example and get that final version out.
billroberts: And you think one more round will do it?
roba: Not a lot of comments and this is going out as experimental so, yes
roba: But noting that others might want to run with this, we may need to do more
phila: Rambles on about responsibility to 'get it right' as far as possible
roba: QB4St doesn't cover as much as it could
… Some sort of standard vocab for this is useful
… It can do more
phila: Then I look forward to those being added in future
billroberts: Anything you want us to cover next week in Delft?
roba: Nothing major
… It's being referenced as a future work piece in the BP doc
roba: One slight question - I have yet to merge Bill's pull request as I'm not sure wwhether you had included that in your PR.
billroberts: I didn't check that there were(n't) other PRs
roba: May need clarification... it's bad form to accept your own pull requests?
phila: Don't get too hung up on it :-)
billroberts: I'm happy to take a look if you're going to feel more comfortable.
roba: That would be very helpful
billroberts: Then I think our session in Delft is going to principally be the vote on CoevrageJSON and an update on the status of the others. And how we link up with the BP doc.
billroberts: I'm assuming that the Australian contingent won't be in Delft
billroberts: I requested a morning slot so you can dial in
kerry: There's the link to the agenda. It's at a reasonable time
… SSN is also keen for a morning slot too
… We can't get into the building until 08:45
… The question, if SSN starts at 08:45, then breaks for coverage and then goes back to SSN - might be easier for Armin
billroberts: I'll be there all day. Any time is OK for me
… It's about how late the coverage session can be. It's about how late it can be for Australia
… Later is OK for me but I'm local
kerry: The earlier slot might work too
<Zakim> phila, you wanted to talk about 08:45
phila: Doors open at 08:45 doesn't mean starting at that time
billroberts: It makes more sense to do SSN in a block
kerry: 9 or 12 sound reasonable
roba: Like that coverage will be short and sweet
kerry: That's a reason to put it first
billroberts: So let's make coverage 9-10 (might not even be that long) and then SSN
kerry: Fine by me
sam_toyer: And OK with me
billroberts: Kerry noted that Maik R has formally left the WG. As he's an author, is there anything that needs to be done?
… Can he be an ex-member and still an author
<kerry> and affiliation stays as is at Reading?