W3C home > Mailing lists > Public > public-poiwg@w3.org > May 2011

Re: POI Core strawman

From: Thomas Wrobel <darkflame@gmail.com>
Date: Thu, 5 May 2011 01:30:46 +0200
Message-ID: <BANLkTi==0fJDpvPB9Rr7ko-8eQzO7dYKYg@mail.gmail.com>
To: roBman@mob-labs.com
Cc: "public-poiwg@w3.org" <public-poiwg@w3.org>
Agreed it should be one of a few options.
Sometimes inline data makes more sense, sometimes remote (normaly
based on filesize)

As far as 3D AR community goes, theres a fair push for x3d format.
This could be delivered as a remote binary file, or simple things
could be done inline as a type of XML. I dont think its necessary for
us to pick and choose formats, but it does illustrate both remote and
local can be a nice flexibility to have.

~~~~~~
Reviews of anything, by anyone;
www.rateoholic.co.uk
Please try out my new site and give feedback :)



On 5 May 2011 01:03, Rob Manson <roBman@mob-labs.com> wrote:
> Although I don't think it should mandate an external URL.
>
> Inline definitions and perhaps even cid: style local references would be
> good options too.
>
> But at least all 3 cases should be explored and URLs should definitely
> be supported.
>
>
> roBman
>
>
> On Wed, 2011-05-04 at 09:36 +1000, Rob Manson wrote:
>> Yep...I think a URL is perfect.
>>
>> Then using something like a rel link would allow multiple
>> representations a bit like CSS and RDF.
>>
>> Then client apps could choose which one to use or use some form of
>> progressive enhancement.
>>
>>
>> roBman
>>
>>
>> On Wed, 2011-05-04 at 00:39 +0200, Thomas Wrobel wrote:
>> > "I see a POI as a useful tool for linking or binding "things" to "locations"."
>> >
>> > Indeed. I see that as its primary goal really - once again I like the
>> > term physical hyperlinks (or plinks!).
>> >
>> > I picture it basically as;
>> > Real Life <POI> Data
>> >
>> > The POI containing the information to locate something in the real
>> > world and (optionally) a link to a bit of data to place there..
>> >
>> > At least when it comes to AR this is the main use of POIs.
>> > But does not this linking require a data element ? or at least, some
>> > way to link to the (remore source) data being positioned?
>> > Not sure if data element is the best name for it though.
>> > As you say it opens a lot of other things to be added based on that,
>> > but at its simplist its just url to...something else...no ?
>> >
>> >
>> > On 3 May 2011 07:56, Rob Manson <roBman@mob-labs.com> wrote:
>> > > I think this is a good point Thomas...and kind of the opposite of the
>> > > recent discussion about "should a POI require a Location?".
>> > >
>> > > It's really a question about "should a POI allow MORE than a Location?"
>> > >
>> > > Again I would think the the Point in POI suggest no.  I see a POI as a
>> > > useful tool for linking or binding "things" to "locations".  Making
>> > > these other geometries as well as other media just "things" that are
>> > > "related" to the POI makes it much more flexible and open ended.
>> > >
>> > > Otherwise are we really just re-inventing or revisiting the type of work
>> > > started in geojson[1], etc?
>> > >
>> > > If yes, then there's a bunch of other things that could be added in
>> > > based on that.
>> > >
>> > >
>> > > roBman
>> > >
>> > > [1] http://wiki.geojson.org/GeoJSON_draft_version_6#Specification
>> > >
>> > >
>> > > On Mon, 2011-05-02 at 21:38 +0200, Thomas Wrobel wrote:
>> > >> Excellent work, that should help get the ball rolling on getting this
>> > >> information pinned down.
>> > >>
>> > >> I'd have to go down the thing element by element, but I see a few
>> > >> things that I would change.
>> > >>
>> > >> * If you use an object to define a georeferance , for example, you'd
>> > >> still need the centre point defined in some other way as well....an
>> > >> object could be used to represent a relative volume, but it doesn't
>> > >> specify a location by itself. So I think that should be mentioned.
>> > >> Also, it should be clear that the definition of an area by object is
>> > >> quite different to specifying a 3d object to place at a POI (which is
>> > >> of course the typical AR use-case).
>> > >> Essentially I think its easy at the moment to get location information
>> > >> and the data which it places mixed up.  Geo-reference should be "where
>> > >> you put something" but what you put there should be specified
>> > >> elsewhere in the POI.
>> > >>
>> > >> * Unless I'm missing something, I think the document also needs a
>> > >> element for this data to be positioned...I'd vote for simple a "data"
>> > >> element that takes a URL and a MIME type - letting people position any
>> > >> arbitrary data or document with the POI. For AR use this would be the
>> > >> link to the 3D model, but it would also make it easy to link a
>> > >> location to a webpage, or even a sound file. (making audio tours very
>> > >> easy to make).
>> > >> significant/worthy to be discussed?
>> > >> Either way, great start to work from.
>> > >>
>> > >>
>> > >> -Thomas Wrobel
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >>
>> > >> I had a (very rough) Time element worked out
>> > >> On 2 May 2011 19:59, Matt Womer <mdw@w3.org> wrote:
>> > >> > Hi all,
>> > >> >
>> > >> > In the interest of getting us closer to publishing a First Public Working Draft, I've created a skeleton editor's draft that is based on materials in the wiki [1], the mailing list, the discussions we have had during calls and face-to-face meetings, and some stuff (the XML syntax in particular) that I just made up as I went along.  There are lots of parts that are incomplete, insufficient, and probably extremely dangerous, but I think it should provide us a good framework for discussions going forward.
>> > >> >
>> > >> > Here it is, enjoy:
>> > >> >        http://www.w3.org/2010/POI/documents/Core/POI%20Core%20Draft.html
>> > >> >
>> > >> > I would ask that as we discuss this, please create new email threads for significant issues rather than just responding to this message.  In the interest of not needlessly spamming the list, please send typos, editorial nits, etc to me directly.  Please also cite the date on the top of the document as well.
>> > >> >
>> > >> > Thanks,
>> > >> >
>> > >> > -Matt Womer
>> > >> > Editor pro tempore
>> > >> >
>> > >> > [1] In particular: http://www.w3.org/2010/POI/wiki/Core/Draft http://www.w3.org/2010/POI/wiki/Data_Model and http://www.w3.org/2010/POI/wiki/
>> > >> >
>> > >> >
>> > >> >
>> > >>
>> > >>
>> > >
>> > >
>> > >
>> >
>> >
>>
>>
>>
>
>
>
Received on Wednesday, 4 May 2011 23:31:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 4 May 2011 23:31:15 GMT