Spatial Data on the Web Working Group Teleconference

<trackbot> Date: 17 June 2015

<eparsons> scribe: linda

<scribe> scribe: Linda

<phila> scribeNick: Linda

Approve Minutes

<eparsons> http://www.w3.org/2015/06/10-sdw-minutes

eparsons: opens meeting
... first item: approve last weeks minutes

<phila> Not present so can't say

<billroberts> +1

<eparsons> PROPOSED: Accept last weeks minutes

<Frans> +1

<ahaller2> +1

<kerry> +1

<MattPerry> +1

<eparsons> +1

<Alejandro_Llaves> +1

+1

<lewismc> +0

<eparsons> RESOLVED: Accept last week's minutes

Patent Call

eparsons: next item: patent call

<eparsons> https://www.w3.org/2015/spatial/wiki/Patent_Call

Best Practice principles

eparsons: main topic today is the principles
... announcing the editors of the best practice document:
... one of whom is on the call: Lewis

<eparsons> Welcome Editors,

lewismc: introduces himself

<eparsons> Lewis McGibbney (NASA)

<eparsons> Payam Barnaghi (University of Surrey)

lewismc: looking forward to it, thanks for the opportunity

<eparsons> Jeremy Tandy (Met Office)

eparsons: other two are not here today, Payam and Jeremy. So we have 3 editors.
... will be an important piece of work

<eparsons> https://www.w3.org/2015/spatial/wiki/BP_Principles

eparsons: proposal: work throught the BP principles together during this call. Numbered them for easy of reference.
... principles are to create the context in which the BP is set and the audience.

<ChrisLittle> +1 agreed

<lewismc> +1

<AndreaPerego> +1

eparsons: important to agree on the idea that the audience of the BP doc is going to be as broad as possible.

<lewismc> Absolutely

<Frans> +1

<Alejandro_Llaves> +1

<billroberts> +1 I agree with that

<ahaller2> +1

+1

<MattPerry> +1

<LarsG> +1

<AndreaPerego> +1

<kerry> +1 -- but remember scope of 5* linked data

Josh: would be useful to [...] (didn't hear)

phila: I hope that web people could read this without any spatial literacy. It's about bridging that gap.

<joshlieberman> Best practices should address a broad audience, but assert / communicate a reasonable level of geospatial literacy.

thx josh

<AndreaPerego> +1 to PhilA

scribe: terminology will be important. BP should help people, not turn them away.
... if possible, without turning away those who are GI experts.

<ChrisLittle> asking for earth, moon, mars etc ;-)

<Frans> +1 to PhilA

joshlieberman: example of spatial literacy:
... web community and spatial community define coordinates differently.

<AndreaPerego> +1 again to PhilA about the terminology issue. One example is "feature", that is a notion difficult to understand without a geospatial background

eparsons: I agree, there are more examples of confusions like this. So we need to make simplifications and give the minimal explanation necessary.

<joshlieberman> We can assume little "prior" spatial literacy, but let's improve on that.

<Zakim> phila, you wanted to suggest a glossary might be helpful

joshlieberman: the BP could assume little prior knowledge, but should provide the necessary knowledge.

phila: e.g. things like coordinate ref system should be defined somewhere because a lot of people do not know about this.
... although it is not our job to educate the world

Frans: simplicity and complexity don't have to be exclusive. You could have simplicity and give the ability to drill down into more detail and complexity.

kerry: In order to attain simplicity we have to give up completeness.

eparsons: agreed. We'll have to ignore the edge cases.

<joshlieberman> So, include spatial omelets, but maybe leave out souffle's

eparsons: or is there some middle way?

<phila> +1 to omlettes

eparsons: I like Josh's point
... proposes that we recognize a mainstream, web audience
... not necessary of professionals

<kerry> +1

<Alejandro_Llaves> +1

billroberts: if we don't aim at simplicity and mainstream audience the risk is that we are just addressing a community that already knows this.
... therefor simplicity is more important than completeness.

<AndreaPerego> +1 to BillRoberts

Frans: could we divide the audience into consumers and publishers?

<ChrisLittle> +1 consumers versus publishers

<Zakim> phila, you wanted to talk about the DWBP experience

eparsons: we may well want to come back to talking about that division. Maybe divide the doc intos sections for those audiences.

phila: discussed this also in data on the web group

<joshlieberman> Somewhat uncertain about this -- everyone with a cellphone is a spatial data producer.

phila: 1st audience was producers, 2nd was consumers

<eparsons> PROPOSED: Audience is the broad web community of non geo-experts

<Frans> +1

<phila> I would say 'an audience is...'

<Zakim> kerry, you wanted to pseak on non-experts

<kerry> https://www.w3.org/2015/spatial/wiki/Working_Use_Cases#Landsat_data_services_.28Best_Practice.2C_Time.2C_Coverage.29

kerry: we have a use case which describes the audience

eparsons: this is from the consumer point of view

<phila> http://www.w3.org/2015/Talks/0612_phila_agile/#(17)

phila: I presented this to ISO TC211 last week
... there must be a happy medium, in this group there is spatial expertise available, what we should do is bring this knowledge to the world

eparsons: So in general we agree on this principle of meeting the needs of this broad audience

<AndreaPerego> +1

eparsons: this leads to the question of approach
... linked data is mentioned in a couple of the principles and underlies the W3C way of publishing data, but not very much adopted in OGC world

<joshlieberman> Does the GeoJSON / Leaflet-using web developer need to know that leaflet is projecting the GeoJSON coordinates into Web Mercator to work with a basemap?

eparsons: so do we suggest this broad audience uses a linked data approach?

billroberts: it is a good basis, just starting with giving everything a url. But lets not restrict ourselves to linked data.
... also do things with JSON, as this is simpler than linked data - or is believed so.

eparsons: requirements are discoverability and linkability. Historically spatial data does not score well on these.

<kerry> but u can do rdf and json at the same time; http://www.w3.org/TR/json-ld/

eparsons: linked data is a possible solution to both but might not be the only one.

billroberts: yes a good solution, but don't advocate as the only one if you want to reach a broad audience.

Frans: a lot of reqs point in the direction of linked data.
... maybe we don't have to be explicit about linked data, it follows from the reqs.
... there's also a risk that we don't critically assess.

ChrisLittle: linked data isused a lot in catalogs, pointing to the data.

<joshlieberman> I propose that we use linked data where appropriate -- e.g. fine-grained data distribution and discovery through mereo-topological relationships. Bounding-box and dimensional discovery doesn't really benefit there.

ChrisLittle: my concern is that in our community we have a lot of 4D data.

AndreaPerego: billroberts made a good point highlighting that linked data can be difficult.
... important requirement is to be web-friendly.

<kerry> note from our charter: The scope of the Spatial Data on the Web Working Group, SDWWG, is Web technologies as they may be applied to location. Where relevant, it will promote Linked Data using the 5 Stars of Linked Data paradigm, but this will not be to the exclusion of other technologies.

eparsons: kerry can you remind us what to do to get 5 stars?

kerry: not difficult: use linked data standards, and use urls.

<Frans> 5 star Open Data: http://5stardata.info/

<Zakim> phila, you wanted to talk about URIs and USB sticks

kerry: we aren't constrained to do linked data but I hope we can promote it.

phila: fundamentals: URIs and HTTP
... don't use the web as a glorified USB stick
... or only publish metadata, not the data on the web

<Frans> We should get those fundamentals in the BP principles.

<joshlieberman> Increasingly clear that the path from linked data to Web-friendliness is URI dereferencing practice.

billroberts: also help people from the broad community to find the right tools

eparsons: we've got some level of agreement to use or inspire to use linked data

<AndreaPerego> +1 to BillRoberts. Which also includes the availability of data in mainstream formats, whenever possible / feasible.

<Frans> Who will be in charge of t-shirts?

eparsons: we should print some tshirts with phila and josh quotes about usb sticks and souffles.
... time to move on to the next topic

Proposed timeline

eparsons: which is the timeline.
... we should discuss reorganising the time frame.
... the bp is the next major piece of work.

<phila> End October

eparsons: we hope to get a first pwd at the time of the TPAC, end of october.

<phila> Mon-Tue 26-27 October

eparsons: although during the summer work will slow down a bit
... the other deliverables, the SSN work, coverages, we could potentially start in september and work parallel.
... so the BP document doesn't have to be complete before starting on the rest. What is your opinion?

Frans: in the original planning bp deliverable had the same deadline as Time ontology. What was the original thinking about dependencies between deliverables?

eparsons: wasn't involved in the timeframe. Phil?

<lewismc> Where is the original timeline/rodmap for the WG? Can someone provide a URL?

phila: we were trying to get this done in two years. We didn't think so much about dependencies as about priorities.

<kerry> timeline is on charter http://www.w3.org/2015/spatial/charter

phila: driving force was deadline end of 2016.

<lewismc> thnx kerry

eparsons: other thoughs?

kerry: I support this, but I'm not keen on parallel meetings.
... try to keep the meetings in a single track, maybe alternate every 2 weeks.

<AndreaPerego> +1 to Kerry

eparsons: I'm not getting the impression that our proposed timeframe doesn't make sense, so we'll work some more on it.

TPAC Registration

<phila> TPAC

phila: W3C big annual get together, the W3C equivalent of the OGC TC meeting.
... a lot of working groups meet there for 2 full days. E.g. the geolocation working group and the Web of Things interest group.
... everyone's welcome to come
... the wednesday is plenary day.
... this year it's in Sapporo, Japan. Our group is meeting monday and tuesday of that week.
... there is a charge for attending, you can register now, please do so.
... wednesday is recommended!

eparsons: will be there monday till wednesday

<joshlieberman> Would be lovely to afford it...

fee is about 85 dollars per day

phila: second 'holiday' is OGC TC meeting in the week of the 14th of september
... the idea is not to have a formal f2f there, but to have this wednesday call take place from the TC.

ChrisLittle: OGC is using gotomeeting instead of webex.

<billroberts> sorry got to go - thanks, bye

phila: doesn't matter

ANOB

eparsons: any other business?

<AndreaPerego> I won't be in Sapporo, but I might be able to be in Nottingham for the OGC TC.

kerry: we also have the idea to do the call from the OGC TC meeting in Sidney in december.

<lewismc> Thanks folks

eparsons: that's it for this week

<lewismc> bb

<AndreaPerego> Thanks, bye

Bye all!

<LarsG> Thanks

<Alejandro_Llaves> thanks, bye!

<kerry> bye!

<MattPerry> quit