Spatial Data on the Web BP Sub Group Teleconference

<jtandy> agenda https://www.w3.org/2015/spatial/wiki/Meetings:BP-Telecon20170118

I can scribe now

<ClemensPortele> jtandy: Probably no point in approving last meeting minutes from November 09

<ClemensPortele> (No objections)

<AndreaPerego> Same here.

Discuss progress based on the Sprint Plan

Discussing the progress

<AndreaPerego> Patent call: https://www.w3.org/2015/spatial/wiki/Patent_Call

<jtandy> https://www.w3.org/2015/spatial/wiki/Detailed_planning_BP_document

discussing the sprint progress

<jtandy> ACTION: 242 to Restructure the document to move the summary to the top and remove the template. [recorded in http://www.w3.org/2017/01/18-sdwbp-minutes.html#action01]

<trackbot> Error finding '242'. You can review and register nicknames at <http://www.w3.org/2015/spatial/track/users>.

Linda: this action is closed.

<scribe> ACTION: 232, 234 to related to BP4 [recorded in http://www.w3.org/2017/01/18-sdwbp-minutes.html#action02]

<trackbot> Error finding '232,'. You can review and register nicknames at <http://www.w3.org/2015/spatial/track/users>.

ClemensPortele: most of the work is done; this is related to making the data indexable and searchable. ClemensPortele is working on adding examples from schema.org

updates will be added this week

jtandy: making the data indexable refers to create human readable pages and also improve the ranking of the pages - by using schema.org- and also by using the way that the crawlers work.

+q

ClemensPortele: providing HTML pages from datasets; and providing metadata and structured data for the datasets
... there are other related matters; e.g. spatial data have identifiers that will be included

jtandy: we also discussed including links in the HTML pages and multiple formats

ClemensPortele each dataset will have a landing (HTML) page.

+q

Payam_ machine triggered search and metadata should be also considered.

ClemensPortele: schema.org metadata is way to address this; this Best practice BP4 is mainly for search engines and HTML should be the key landing point

ChrisLittle_: to make this work, you need to provide some indication of the structure of the data

<ClemensPortele> https://developers.google.com/search/docs/guides/intro-structured-data

ClemensPortele: if you provide your data structured data based on schema.org (link above)

if we have a webpage about a building this will also include information such as location (in structured form) that machines can extract

ChrisLittle_: is concerned with big data and how you get to that structure once you have very large datasets
... we may need to add some explanation to clarify how the proposed best practices should be scaled to big data

jtandy: data on the web best practice has a best practice related to this
... difference between dataset and distribution: will this be also part of the BP4?

ClemensPortele: no; maybe this is related to a different BP

AndreaPerego: the current BP description focuses on optimising the publication for human users and at the same optimising the metadata related to the datasets; this won't exclude providing the data in other formats

<AndreaPerego> Schema.org 4 Datasets: https://www.w3.org/community/schemaorg4datasets/

AndreaPerego: we should provide the data in both machine readable and human readable format (from data on the web best practices)
... we should also consider including requirements that are not currently addressed by schema.org

Next item: Re-writing Section 7

Ed has provided some new text and Linda has merged it - editor's draft

jtandy encouraging everyone to read and comment on this section

ByronCinNZ: talks about BP3
... more that CRS was about the data that affects than- ByronCinNZ has updated this. will do an accuracy check and will make the updates visible soon.
... asking if we have discussed accuracy and precision in the text
... is substantially re-written thhas BP and has provided some examples
... was this the right direction? is bringing data discussion beyond the scope of this group?

jtandy: one of the editors can review and decide on how to integrate the updates

<jtandy> [amendment: suggest that @ByronCinNZ submits a PR for the WG to review in the ED; editors will process the PR]

ChrisLittle_: discusses the importance of having a statement on accuracy and precision:
... discusses user or customer oriented CRS; e.g. near to me; this can be future work; lots of work related to this are ad-hoc

<Zakim> jtandy, you wanted to note that BP order will be dealt with in a later iteration

jtandy: BP9 will clarify that there is lack of practice in relation to user or customer oriented CRS

ByronCinNZ: a diagram will be helpful to be included

jtandy: we should make sure there is no IP issues for the diagram/figure-
... swapping the order of the BPs- there is a action and this will be done in (near) future

<Linda> NOTE

<Linda> It is important to understand the difference between precision and accuracy. Seven decimal places of a latitude degree corresponds to about one centimeter. Whatever the precision of the specified coordinates, the accuracy of positioning on the actual earth's surface using WGS84 will only approach about a meter horizontally and may have apparent errors of up to 100 meters vertically, because of assumptions about reference systems, tectonic plate movements and whic

<Linda> h definition of the earth's 'surface' is used.

Linda: there is a note in BP5 on precision and accuracy (text above)

Next items BP14 and BP7 (Jeremy)

jtandy is working on them- he needs contributions form BartvanLeeuwen \

BartvanLeeuwen: discusses his updates
... will provide some updates for BP7

<ChrisLittle_> S/form/from/

AndreaPerego: BP8 and Geometry work
... you seem to be muted
... please try to reconnect
... BP8 and Geometry work
... BP8: we are not making any strong recommendations; we give a number of options but not strong recommendations

<jtandy> see AndreaPerego's email here: https://lists.w3.org/Archives/Public/public-sdw-wg/2017Jan/0081.html

AndreaPerego: is this sufficient or should we provide more detailed recommendations and examples
... preferred geometry representations; limitations of the current methods and other methods such as geo-sparql
... seems the group is undecided about the encodings
... the 2nd part is on how to publish geometry on the web
... wether you publish the geometry associated with the spatial thing or separately
... we have examples for each of these cases
... we have to define and clarify the use-case to make this more practical
... how to make the geometry re-usable (but geometry could have different levels of granularity and accuracy)
... however, authoritative geometry is fixed and this should be one of the use-cases

and we need to clarify depending what a user wants to do, s/he can select one of the approaches

we also need to provide guidance on the requirements and the representations and what should be included in them; eg. some only provide geometry in RDF; but some others also provide information about serialisation;

we seem to be a bit unclear about providing guidance about how to provide the geometry

jtandy we haven't make any conclusions yet; but AndreaPerego discussions will help to provide some clarity on this topic.

jtandy: referring to the a case that RDF is separated from RDF; due to efficiency issues

jtandy balancing the performance against what trying to achieve is also important

AndreaPerego: If data is to large to be included in RDF, you can keep this out but provide link (via RDF) where the serialisation of the data is available

<Zakim> jtandy, you wanted to ask about using GML for geometry

<Zakim> ClemensPortele, you wanted to talk about GML, too

ClemensPortele: talks about GML

<jtandy> [GML is not webby? ... I think that there are javascript libraries that can parse GML]

<jtandy> [GML is also pretty much our only option for 3D data]

ClemensPortele if someone working in an XML environment, GML will be the choice

ClemensPortele: there are other cases that GML is a better choice;

<Linda> jtandy andrea can you include me friday? i can share experiences from testbed regarding geometries inside/outside triple stores & performance

ClemensPortele: we should provide guidance on the choices and criteria for these

<jtandy> ClemensPortele: GML is our only option for describing solids at present (3D)

AndreaPerego: talks about GML and issues that it can address - we shouldn't be restrictive; people should be able to publish in their own format; but we should give them a (set of) recommended format(s)

ClemensPortele: will continue the discussion in the mailing group.

ChrisLittle_: not being webby probably refers to not being suitable for triple store

<BartvanLeeuwen> thx guys

<Linda> bye

<AndreaPerego> Thanks, and bye!

<ChrisLittle_> Bye

bye

<ClemensPortele> bye, thanks!