[Ed takes us through the patent call]
[Jeremy reviews the agenda, asks for any other business]
BartvanLeeuwen: I have a comment on the work that Clemens did. BP11.
jtandy: Duly noted.
jtandy: Good idea to approve minutes from issue resolution meeting
<AndreaPerego> 0 - was not there
<ClemensPortele> +0 (did not attend)
<AndreaPerego> Actually, the minutes of the 2017-04-05 telecon are here: https://www.w3.org/2017/04/05-sdwbp-minutes
<jtandy> WG email thread: https://lists.w3.org/Archives/Public/public-sdw-wg/2017Mar/0268.html
jtandy: Has anybody not seen the reordering proposals?
jtandy: [quickly goes through the options]
… I added an option based on thematic ordering.
… And we have a final one which is based on work done on guiding principles in stats community, which I'm less familiar with.
… I think Ed, you indicated that you were fine with option 5.
Payam: Data on the Web is more or less ordered based on workflow.
jtandy: All of these options are different from the current structure, which follows the structure in the Data on the Web BP document.
Payam: I feel we should spend more time on the content than on the reordering.
jtandy: I hear that, but we resolved to reorder it in a previous meeting.
… In particular, first BP is not seen as the good opening one.
ClemensPortele: Just to mention that my preference for option 4.
… I see the arguments for option 5 and right now I would also agree that option 5 is the best of the options.
… Regarding option 6, this is about aligning with a different structure. That's useful for people that come from that other structure, but it's more useful to follow the order we feel is correct.
eparsons: I think what's nice about option 5 is that it is priority based but also organized around the biggest changes first.
… Further options 6 and 7 are more traditional, what we've already done before. I'm still happy with my vote on Option 5.
jtandy: I think we need to acknowledge that Laurent did a great work on option 6, this is not being ignored, for sure!
ByronCinNZ: Option 5 is fine. I agree we should not lose the categorization done in Option 6.
AndreaPerego: I'm also quite happy with option 5. Only question: the metadata is at the end, and we refer to it in the previous BP, at least in examples.
… So just wondering whether addressing metadata at the end could be an issue.
jtandy: I hear that. I note Linda's report that whatever order she tried, there was no way not to have BPs link to further parts in the specification.
… Option 5 seems the closest we have to a perfect solution so far.
AndreaPerego: Irrespective of the order that we want to use for the reordering, some BP may have been written knowing which BP was before or after in the document, with text that says "previous" or "next". To be checked.
jtandy: Good point.
… I'm adding some notes to my editorial list of actions accordingly.
<jtandy> tbd | Ensure that each best practice when referring to others takes account of the re-ordered sequence of best practices
jtandy: What I've heard is that everybody who's spoken is fine with Option 5. Byron mentioned option 3 as well but still agrees with 5, right?
ByronCinNZ: Yes, option 3 does not have to be exclusive.
jtandy: The section intros will need to be re-assembled, but I think that we're approaching a proposal here.
<jtandy> Proposed: the BP document is restructured according to proposal 5 https://www.w3.org/2015/spatial/wiki/index.php?title=BP_2017_reordering_proposal#Proposal_5:_Thematic_.26_prioritised_ordering
Resolved: the BP document is restructured according to proposal 5 https://www.w3.org/2015/spatial/wiki/index.php?title=BP_2017_reordering_proposal#Proposal_5:_Thematic_.26_prioritised_ordering
Action: Linda to restructure the BP document according to Option 5
<trackbot> Created ACTION-310 - Restructure the bp document according to option 5 [on Linda van den Brink - due 2017-04-19].
eparsons: I picked up the action to have a look at this. Issue popped up from an email asking us to look at a methodology to handle large geometry objects developed at IETF.
… Browsers would struggle with a very large array of vertices. This is going to be an issue with GeoJSON.
… Proposal that Sean made last year did not seem to receive a lot of comments.
… There are methodologies emerging, but it will take time before things settle done.
… My suggestion is to close the issue, noting that this is an emerging area.
jtandy: If I recall, Andrea, you have a bit of outstanding work about providing simplified geometry.
… Would that be a nice addition to it?
AndreaPerego: I think so.
eparsons: I agree, these are basically the same issue.
jtandy: OK, I am updating the BP8 task in the detailed plan
<jtandy> > Andrea with help from Ed | Refer to GeoJSON Text Sequences as just one of the methods to work with large geometry objects; see GitHub Issue 656
eparsons: Essentially, the text in the GitHub issue could be copied-and-pasted.
jtandy: In the work that Kerry has been doing, she proposed something about balancing quality and cost, which I suggested Bill take a look at.
… I also added a change to BP6 to accomodate the "BP change over time" comment.
… In her third point, she talks about referencing EO-QB and CoverageJSON., and UCR in fourth point.
<jtandy> > [VOCAB-DATA-CUBE] provides a generic mechanism to express well-structured data, such as timeseries, in RDF. [EO-QB] and [QB4ST] (developed alongside this best practice Note within the Spatial Data on the Web Working Group) illustrate how [VOCAB-DATA-CUBE] may be used in this way.
jtandy: All I'm doing is referencing the other work that the group has done. I'm not making it a Best Practice, sticking to MAY.
eparsons: I think this is a good approach.
… Kerry was only referring to examples, so perhaps we should be stronger and actually say that these are not best practices
… but they respond to identified challenges.
Payam: Kerry's email says that these are applications of best practices. I agree they are rather examples.
jtandy: I'm a bit confused
eparsons: These are approaches that fill the context of balancing quality and cost, but we're not in a position to say that these are best practices.
… Surely it makes sense that we reference other deliverables that we're doing in the group.
… But there is likely going to be additional work, even after this document is published, on these aspects.
jtandy: Hearing what you said, Ed, we should make the link but identify explicitly that these are not best practices, but do provide mechanisms that may be useful to describe data that vary over time.
jtandy: I'm content with that.
jtandy: Going back to Kerry's point 3, about referencing EO-QB and COV-JSON. I said that section 6 does not reference implementations and I'm happy to leave it that way.
… I have asked Bill to figure out if he can find a way to add a reference to Coverage data models.
jtandy: Are you content with that as an approach?
… So it's all on Bill.
jtandy: I've addressed Kerry's point 4.
eparsons: All in all, it would be nice to link works together, but timing plays against us, and it would have been nice to have more mature specs as well.
… Focusing on the end user is the most important factor.
jtandy: We do have a detailed plan
jtandy: If you're doing a section and you think that you're not going to be able to meet the deadline that you've got, please flag it now.
[some discussion on how great GitHub is]
AndreaPerego: I think I will be able to meet the deadline. I'm currently working offline because changes are related.
… I will put them on my fork on GitHub and then I would like to ask the group to review the relevant bits before I prepare pull requests
… Hopefully next week.
jtandy: Friday 21st should be last day. It may be that we need to slide by a week. Is it sensible for us to agree with that, now?
AndreaPerego: I will drop a mail before completing things to get feedback from the group.
<ClemensPortele> When will Linda do the reordering? Probably after all the other changes?
jtandy: If you keep on fetching gh-pages from w3c repo and rebasing, things should go smoothly.
eparsons: It's probably a good idea to avoid touching the document during the reordering process.
jtandy: That's a good point.
… Why don't you have a discussion with Linda once she gets back from vacation.
BartvanLeeuwen: Clemens did a large restructuring in BP11 and I was wondering whether the part I wanted to have in there is still in there. And I think the answer is "yes", so that's good.
ClemensPortele: Thanks Bart for looking at it. I tried to preserve the key message of the architecture that you've been using.
Payam: [Talking about a "table", scribe confesses being lost here] I can prepare it while in Japan but will need help integrating in the document.
<BartvanLeeuwen> thx jtandy
<ClemensPortele> thanks, bye