- From: Phil Archer <phila@w3.org>
- Date: Wed, 4 May 2016 17:13:59 +0100
- To: SDW WG Public List <public-sdw-wg@w3.org>
Dear all, the minutes of today's BP sub group call are at
https://www.w3.org/2016/05/04-sdwbp-minutes and, of course, as text below.
Spatial Data on the Web WG BP Sub Group Teleconference
04 May 2016
[2]Agenda
[2] https://www.w3.org/2015/spatial/wiki/Meetings:BP-Telecon20160504
See also: [3]IRC log
[3] http://www.w3.org/2016/05/04-sdwbp-irc
Attendees
Present
ScottSimmons, eparsons, frans, phila, billroberts,
jtandy, MattPerry, AndreaPerego
Regrets
Clemens
Chair
Jeremy
Scribe
eparsons
Contents
* [4]Topics
1. [5]patent call
2. [6]Summary of main points from Best Practice sub-group
'virtual meeting', 3-May-2016 Minutes
3. [7]Overview of BP Narrative 2
4. [8]Allocate examples to WG members to complete
5. [9]AOB
* [10]Summary of Action Items
* [11]Summary of Resolutions
__________________________________________________________
<phila> present= phila
<frans> top left?
<scribe> scribe: eparsons
<scribe> topic : last weeks minutes
<jtandy> scribenick: eparsons
<jtandy> [12]http://www.w3.org/2016/05/03-sdwbp-minutes.html
[12] http://www.w3.org/2016/05/03-sdwbp-minutes.html
<jtandy> Proposed: approve minutes of last meeting
<jtandy> +1
<josh> +1
+0
<ScottSimmons> +1
<jtandy> (actually =
[13]http://www.w3.org/2016/04/20-sdwbp-minutes)
[13] http://www.w3.org/2016/04/20-sdwbp-minutes)
<josh> We're being supportive...
<frans> +1
RESOLUTION: approve minutes of last meeting (on 20 April)
patent call
<jtandy> link:
[14]https://www.w3.org/2015/spatial/wiki/Patent_Call
[14] https://www.w3.org/2015/spatial/wiki/Patent_Call
phila Pleased with jtandy good practice of checking irc and
webex participants
phila important for IPR reasions
Summary of main points from Best Practice sub-group 'virtual
meeting', 3-May-2016 Minutes
jtandy josh, linda, phil (decorating) present on call yesterday
- thanks to all
jtandy Seemed to work well, recommended as good way of working
<jtandy>
[15]https://www.w3.org/2015/spatial/wiki/BP_Narrative_2
[15] https://www.w3.org/2015/spatial/wiki/BP_Narrative_2
jtandy flooding story rewrite by jtandy link above
jtandy too techie language perhaps
jtandy Flooding emotive subject - event not prediction of
flooding is relevant only
frans 2 versions really needed ?
jtandy Not tech story with links to details or details mixed
within narrative are the options, tech content is really needed
and important
<jtandy> ack Linda|2
jtandy DRY Don't Repeat Yourself
Linda|2 Domain language was identified yesterday as issue
josh Understandable to web and GI communities to issue
josh Matching terms needed
<josh> Example from yesterday: "spatial join" vs "mashup"
jtandy Key to avoid hydrological terms
jtandy "Real world Thing" is a useful term rather than feature
jtandy Not Happy with thing but best we could come up with
<jtandy> ack Linda|
jtandy Good talk of crowd sourcing - is the audience platform
devs eg. twitter or users of platform api's
jtandy Platform devs temselves eg twitter, facebook not so
different to SDI publishers
jtandy URL's to identify things good in tweets etc..
jtandy eg. what 3 words
<ScottSimmons> sorry - ScottSimmons must leave
<jtandy> "acknowledge other types of platform providers as
audience for BP doc" said Frans
frans Platform devs good audience - Data Storage providers
could also be interested - looking for guidance
jtandy Audience discussion fruitful - SDI providers, twitter,
API builders are all platforms of sorts
josh All web server based technologies are platforms ?
jtandy The web is the one platform to rule them all
jtandy What does Spatial data on the web mean to Small
publishers - links to other resources / large publishers more
complex use of data
jtandy josh ? anything to add
josh similar cases - fine grained links not scalable - large is
scalable
<jtandy> "APIs that add scaleability to that fine grained
access" said Josh
jtandy loves josh
jtandy platonic
jtandy Our examples need to focus on data publication - first
step if not published, cannot be used
eparsons Agree completely
billroberts Happy with focus on data publishers - risk we miss
the step to make it easy for users of data
<josh> The goal of publishing data to enable "better mashups"
means that implementation feedback is crucial.
jtandy Test the usage of data against each BP - "Is the data
easy for users to access"
billroberts Usage patterns need to promote use
jtandy "web as data sharing platform"
Overview of BP Narrative 2
<jtandy>
[16]https://www.w3.org/2015/spatial/wiki/BP_Narrative_2
[16] https://www.w3.org/2015/spatial/wiki/BP_Narrative_2
jtandy Linda|2 has made updates to above
jtandy Key points against 9 steps
<billroberts> what?...wakes up
jtandy First step publish flood prediction data as a coverage
<billroberts> yes I was listening
jtandy Nobody owns theis example - Bill ?
jtandy Second step Admin Areas from SDI source using proxy
layer
jtandy includes CRS, precision etc
jtandy third step - web developer takes coverage data -
converts to vector features
q
<josh> WPS!
eparsons Is that realistic for web dev ?
jtandy Open source dev - might make an app..
josh Might use a restful service to join spatial data
<frans> Spatial joins are often needed
josh spatial join wrong language mashup is more appropriate
jtandy web dev called Helen wants to use her edge finding
algorithm in javascript - so can do that
jtandy Helens vectors are very details - brings in issue of
performance - need to generalise for example ?
jtandy - point taken think about better ways...
jtandy Step Four publish fixed assets, topographical features
not using SDI in this case - normal web tech
jtandy SDI use case but not using SDI tech..
jtandy Step five of interest to billroberts - Figure out number
of people the food covers, CSV file of peoples location
converted to RDF
jtandy RDF data ingested into triple store for SPARQL query
jtandy billroberts Is this your bag ?
billroberts Almost - different tools - but census data as RDF
is relevant
billroberts CSV to RDF then exposed as SPARQL endpoint but
small community
jtandy Capture what billroberts does would be really useful
frans Point 5 realistic ? Census publishers often use more
advanced publishing tools
billroberts 50% agreement with frans - Many do use csv however,
jsonstat another common format, specific api's etc
billroberts Good Step however..
jtandy Realism is aim - BP needs to expose problems
billroberts - Too much detail would be an issue in too real
examples
jtandy Not trying to build flood prevention system, details not
relevant to our use of scenario
frans Something not so based on semantics, weather data might
be better, not as loading with semantic problems in stats
jtandy maybe, pop broken down into groups
jtandy Step six - emergency teams use GIS to create evac plan
jtandy Take GIS analysis product and publish as html, or API
for embedded use on web pages
jtandy eparsons would you work on this
eparsons yes !!
jtandy Step seven data streams - josh to take up
jtandy step eight - social media examples, twitter users
advertising phone charging, Instagram users taking pictures of
floods
josh links to step seven
josh IoT testbed, causing stress with SSN work however
jtandy step nine - most event SAR data cross referenced with
insurance data
Allocate examples to WG members to complete
jtandy - makes sense ?
jtandy - Step 1 coverage data needs someone to help work on..
jtandy - Step 2 clemens
jtandy step 3 - Linda|2
jtandy Step 4 - Andrea
jtandy Step 5 - billroberts
jtandy Step 6 -eparsons
jtandy steps 7 & 8 - josh
jtandy Linda|2 had added to narrative for step 3 good example
AOB
<frans> Thanks to all the BP people!
<josh> bye
<AndreaPerego> bye
Bye
Summary of Action Items
Summary of Resolutions
1. [17]approve minutes of last meeting (on 20 April)
[End of minutes]
__________________________________________________________
Received on Wednesday, 4 May 2016 16:14:11 UTC