W3C

Spatial Data on the Web Working Group Teleconference

31 Aug 2016

Agenda

See also: IRC log

Attendees

Present
roba, DanhLePhuoc, ScottSimmons, BartvanLeeuwen, ByronCinNZ, phila, kerry, frans, ChrisLittle, jtandy, RaulGarciaCastro
Regrets
Rachel, Lars, SimonCox, Andrea, edparsons, Ed
Chair
Kerry, Jeremy
Scribe
kerry, Jeremy Tandy, phila

Contents


<jtandy> http://www.w3.org/2016/08/17-sdw-minutes.html

<ChrisLittle> +1

propose: approve minutes http://www.w3.org/2016/08/17-sdw-minutes.html

approve last meetings minutes

<jtandy> +1

+1

<frans> +1

RESOLUTION: approve minutes http://www.w3.org/2016/08/17-sdw-minutes.html

patent call

<roba> i ma missing from present list again - maybe i beat bot to it, otherwise +1

<jtandy> scribe: Jeremy Tandy

<jtandy> scribenick: jtandy

<ChrisLittle> S/am/am/

f2f https://www.w3.org/2015/spatial/wiki/Meetings:F2F4

<ChrisLittle> S/ma/am

kerry: goal to bring people's attention to the upcoming f2f
... there is a draft agenda
... best so far, [but by no means finished!]
... 8:30 am to 6pm

<BartvanLeeuwen> +1

kerry: is this OK?

<frans> 8:30 is doable

<phila> There is usually more coffee than you can drink

<ChrisLittle> * back in a few minutes

kerry: registration is open from 8am
... summarises ... 8:30 to start
... first thing on the agenda is UCR doc
... this should be the last time we look at this
... is that OK frans ?

frans: there are still things coming on the plate. will there be new versions of the documents before TPAC?

kerry: yes - there was a preference for this - a week before TPAC for stable review.

<phila> scribe: phila

frans: We can have a new version of the UCR before TPAC but it may not be final.

kerry: OK, let's leave it on the agenda
... And that will ..

jtandy: Can I just ask Frans..
... One of the things on the list recently is whether or not we have a req for units of measure and precision and accuracy.
... This is necessary as a minimum for CRS
... from a BP doc perspective, we're not putting anything in until there's a req for it.

<ChrisLittle> * back

jtandy: So are you trying to synthesise those reqs before TPAC.

kerry: Should we move the UCR to after the BP discusssion?

jtandy: I think the BP discussion will take up as much time as it can and more

kerry: If there's anything concrete to add to the agenda, let me know.

<jtandy> https://www.w3.org/2015/spatial/wiki/Meetings:F2F4-best-practice-agenda-scratch-pad

jtandy: I've created a wiki page ^^ trying to get individuals to talk about issues that interest them most.

kerry: If you can work in that time frame, please edit the wiki

jtandy: I notice that you included a vote to release a new version of the BP doc in the agenda...

kerry: How would you like to end?

jtandy: We have a draft doc, we have a list of outstanding changes to make, all allocated to individuals to complete by a given date. Then after TPAC we can do what we need to do and then release after that.

[More jtandy - kerry discussion]

<Zakim> phila, you wanted to ask kerry which agenda we're looking at as https://www.w3.org/2015/spatial/wiki/Meetings:F2F4 seems different

-> https://www.w3.org/2015/spatial/wiki/Meetings:F2F4

phila: Confused

kerry: Updating the agenda again...

frans: To be clear, there won't be a new PWD of the BP doc before TPAC, but will there be a stable version to review before then

jtandy: of the BP doc? There should be a reasonably stable version in the week up to TPAC.
... We'll highlight where there are gaps.
... Linda is now away for a week and a bit
... I'm overlapping... so we may be a little uncoordinated up to TPAC.

Tuesday

kerry: Bill has asked for time on coverages. we'll start first thing
... Up until morning tea. And then the rest of the day is SSN. The SSN people need some plenary time.
... There will be sub topics
... Important for us to come back to the main group
... And then the plan fort the last 2 hours - we've been approached by the Web of Things WG
... They're spawning a new group who want to talk to us about a vocab for a Web of Things
... We were aiming for a session on Wednesday but they're not available.
... So I think that WoT session will be interesting beyind the SSN people. Lots of spatial data things to think about.

-> www.w3.org/blog/news/archives/5751 Sensor API stuff

<ChrisLittle> +q

<kerry> phil asks for review that paper

phila: Plugs the generic sensor API seeking review

kerry: Danh - could you take a look at that, please?

<joshlieberman> It would be helpful for those of us conflicted in Orlando to make sure the TPAC session notes are fairly complete and promptly available.

DanhLePhuoc_: OK, I can take a look

ChrisLittle: Can DanhLePhuoc_ please let me know about his review directly? So I can report back to OGC?
... Lots of relevant stuff at OGC on this

<ScottSimmons> +1 to Chris

DanhLePhuoc_: I see several things in common. So yes.
... I'll look at that.

kerry: I have to say - Chris, your suggestion is very sincere - everyone would agree with you but it's a very big space.
... OneM2M is based on SSN I believe.
... There are lots^n of groups working in that space.
... There was a big workshop in Europe last year that mapped that.

ChrisLittle: Then maybe we should be in the business of mapping the landscape.

<joshlieberman> Perhaps the best we can hope for is to try to keep up on how all of these API's may be related to each other. OGC are still dealing with the overlap between STA and SOS...

ChrisLittle: Can we look at common approaches

ScottSimmons: I agree

<frans> It seems to me that the sooner convergence attempts are started, the better.

ScottSimmons: That was news to me too. We should try and get the groups talking
... I have a rough idea where the differences anda similarities lie

<Zakim> phila, you wanted to make a suggestion to Denise

kerry: We have Robin ? who is working with Steve Liang who was coming to SSN meetings. He's never said much

phila: He's never spoken at all as far as I know

kerry: And the API people - Linda and I went to their meeting in Japan last year
... Didn't feel it was on target for us

phila: They're only concerned with browser APIs, not what's behind it

<kerry> q/

phila: We should check that we should or shouldn't be talking

DanhLePhuoc_: DO they see spatial data as relevant to sensor data? Even within W3C we had ?? context
... then we... [very faint]
... I don't know how we can collaborate

<kerry> danh: spatial data also relevant to sensors, also [something] and snesor api and web of things and other iot consortiums... I don't know how we can collaborate

<ChrisLittle> S/sensor/sensor/

<ChrisLittle> S/snesor/sensor/

kerry: So we're waiting for a bit of filling in from SSN and BP for the TPAC agenda.

<DanhLePhuoc_> how to position ourselfs and collaborate with other groups, Device Context, Sensor API, WoT, ETSI M2M, oneM2M which have relevant APIs and data model

<ScottSimmons> ScottSimmons has just sent a message to the OGC SensorThings email reflector to request participation/support for engagement with generic sensor API group

Updates on the UCR doc

BP update

frans: I haven't prepared anything. There are many different topics. I'd like to ask people to pay attention to the mailing list ...

Best Practices

jtandy: Linda and I have been busy trying to press on with the doc. I think she's done great work in boiling it down. We used to have 30, now we have 15.
... Quality not quantity.
... Linda has also been working to align them with DWBP
... Some of the things that are left over... we think we want a section on CRS which Payam is leading.
... We anticipate a section in the intro on CRS and a BP around how to choose it. That will need to apply to non-geographic CRS ones.
... We need a decision tree on choosing your vocabs and formats
... How to we channel people to the simplest format that does the job.
... we have the Newhaven flooding scenario.
... My job is to clarify what my actors are doing when publishing admin boundaries, flood extent, using social media...
... And then I need real world examples.
... This won't be done before TPAC
... Really want to have some real world examples and data
... Maybe that not all the examples will relate to flooding.
... Not all our data will be geolocated.
... the data examples won't all be in the same place.

<BartvanLeeuwen> looking for unmute button

jtandy: I think Bart is offering some geolocated examples in a test bed?

BartvanLeeuwen: We're not talking about a test bed but we are pursuing the stuff I demod earlier.
... We'll be presenting that the day after I've been to TPAC.
... We want to say it's easy to add this, to expose data etc.

jtandy: The examples being geographically distributed is probably correct then

<Zakim> phila, you wanted to talk i18n

-> https://www.w3.org/TR/dwbp/#LocaleParametersMetadata DWBP CR

phila: Talked about the slight change to locale neutral advice in DWBP

kerry: When we were doing SSN the first time, we were working with people doing stuff on the other side of the channel - Southampton/Solent
... I think that was public and had lots of data

<ChrisLittle> +q

jtandy: There will be limited time between now and TPAC to get the data into the doc, but I do want to create place holders
... If you can find some references, so much the better.

ChrisLittle: I admit I haven't read you scenario. Does it say anything about tides?
... So it's the excessive rainfall AND the tides that matter.
... It's a flourishing coastal city in NL... We don't go into causes of the rising water.

jtandy: There are some e-mails that ask for input on nspecific BPs
... 5, 6, 9, 16, 17
... I need help please.
... One of those, 6, is about the use of identifiers. I'm looking for people to indicate support indirect identification. It talks about using the identifier of a mailbox to identify the person whose mailbox it is.
... I'm looking for consensus to emerge.
... The sooner I can get those issues solved, the sooner I can get the doc into shape.

<joshlieberman> Jeremy, was my response clear?

joshlieberman: There is this question, do we have this HTTP Range 14 problem?
... The assertion is that it doesn't because we're talking about feature data since if were clear that we're talking about feature data, then that's clear
... There is some legitimate ambiguity that not everyone cares about.
... It's an inherent property of feature data.
... If we need to disambiguate which real world feature we're talking about. This feature representation may not be the same as that, although they may be referring to the same real world featuyre
... In many cases, people are referring to the same real world thing, so there are occasions where we may need to disambiguate. Otherwise I think the indoirect ids worl well for spatial data.

jtandy: I'll re-read what you said and then come back to you if needs be

<ChrisLittle> S/work/work/

roba: I won't touch that topic. My concern... I see a degree of inconsistency between subgroups around whether a requirement must be in the UCR before it can be dealt with.

<ChrisLittle> S/indoirect/indirect/

roba: Can anyone clear state what the relationship is?

frans: The way you said it is the way the UCR is set up. It specifies reqs for Spatial Data in the Web. Sub groups can work with other sources thayt have additional requirements.
... Some topics go beyond spatial data on the web, such as time.
... So I think it makes sense to have some exceptions.
... If a requirement just comes from generally doing things right then I don't think we need to include it.

roba: So it's not 100% consistent with the idea that a req must be in the UCR

kerry: I've not heard that about Time or SSN

roba: For example, the SSN group asked me to see if the UCR was complete

SSN

kerry: We wanted to make sure that SSN was meeting its reqs
... we resolved that in our last SSN meeting

<frans> I am behind on a few topics, I have to admit.

kerry: I don't think SSN needs to be entirely constrained by the UCR, but we'd be mad not to cover all the reqs in the UCR

<roba> ok - happy that we are saying that we must have UCR ref for every requirement...

frans: Two new requirements are coming, so we're still accepting new ones.

roba: OK, I;m happy if everyone accepts that we don't *have* to have a UCR Req

<jtandy> agree with rob's assessment

kerry: Nor do we have to deliver all the reqs, but we have to do a mapping and if we don't cover it, say why.

<roba> thanks phil

kerry: On Coverages... we have the first template for a doc around data cube...

<BartvanLeeuwen> thx bye

<ChrisLittle> Bye

<RaulGarciaCastro> Bye!

<jtandy> bye

<kerry> chair: kerry

Summary of Action Items

Summary of Resolutions

  1. approve minutes http://www.w3.org/2016/08/17-sdw-minutes.html
[End of minutes]