We've been requested by Phil Archer and Christine Perrey to -please- put
together some contributions for the W3C AR meeting in Barcelona next
week. I Know Damon Hernandez is going to summarize the outcomes of the
X3d AR workshop at ARE, Thomas and perhaps Tish will submit something
about ARwave and XMPP and I hope Blair and Alex are going to send
something about their shiny new Kamra browser.
That leaves the rest of us. I' plan to write up a short issues paper,
over the weekend about the entire stack that I'll share here for
comments before sending it tho the W3C. I hope Anselm and Sophia will
write something about open geo and FOSS for AR, (?) and maybe we can
convice Dan Brickley to contibute something about AR semantics and maybe
RESTful services. (?)
That still leaves about 70 people signed up for this list, who have, so
far been mostly silent. Let's hope some of you have something to add to
the discussions. this is a great opportunity to make an major impact on
the future of open AR software and data.
Cheers!
Mike
In my own time at the moment I'm building a simple (and open source)
AR game level authoring tool - it's the sweet spot where I feel the
need is greatest. I'm finding a need to have a formal description
grammar that describes not just what is in the world, but what the
relationships between things are. Also I am finding a need to formally
express properties such as time, duration, sensor radius, sensor
filter conditions and the like.
--
@anselm 415 215 4856 http://twitter.com/anselm
Cheers
Phil.
--
Phil Archer
http://philarcher.org/
@philarcher1
i-sieve Technologies | W3C
Sentiment Analysis Beyond Impressions | Mobile Web Initiative
http://i-sieve.com | http://www.w3.org/Mobile
Ah, now my guilt kicks in! I *really* wanted to get a position paper done.
Here's my thinking in a scribbly nutshell, maybe it fits with others' thinking?
* AR is necessarily a hodge-podge of fun things, bridging fields of
expertise and standards; there will never be a single suite of 'AR'
standards...
* AR connects with SemWeb when it comes to building apps that require
acting upon diverse descriptions; even the coolest swooshy 3d display
can only usefully show you a handful of items. Appropriate filtering
and ranking requires a richer schema than "short bit of text + long
bit of text + category + logo" and what we call AR is just the tip of
a larger iceberg.
* W3C should pay some attention to QR Code standards and suchlike
(also imho audio encodings); they already exist, but best practice
guidelines for URIs is important. Imagine if each shop / point of
interest had an array of 5 different scannable codes on the front door
or in magazines and flyers - not going to work. A single consensus on
how to encode your homepage URL in scannable form, plus some work on
what markup to embed in that homepage to describe a 'point of
interest' richly is needed.
* more effort linking with especially egov open data efforts. A lot
of AR is about cities in particular, due to the massive concentration
of people and information. AR apps can provide the shiny demos that
help civil servants open up data. And the data doesn't need to be in
some new "AR" format; XML, CSV, RDF etc are just fine; it is a mistake
if we fixate on the 3d and geo specs as the main requirement that AR
has from the standards world. Not that the 3d stuff isn't important
(but then so is oauth, qr, etc too)
* we need some strong demos showing content portability between AR
environments. In a year or so the young creative AR startups will be
in a squeeze: oppressed from one side by amazing demos from the giants
(microsoft, google, apple) and richer builtins on the mobile
platforms, and from the other by increasingly feasibility of doing
basic swooshy camera overlay stuff in js via HTML5 + APIs, ie. richer
builtins on the Web-as-platform. In particular, social AR apps will
remain somewhat stunted all the time they assume we're all running the
same software --- I want to annotate the world for all my friends
regardless of the phone or software they're currently using, and to
have that effort still be worthwhile in 5+ years time.
Concrete things I'd like to see in the coming weeks: some AR-motivated
extensions to Facebook's http://opengraphprotocol.org/ that make it
better support some practical "make AR useful" scenarios (what movies
are on? (that I might like? that I haven't seen...) where's a *vegan*
restaurant near here that's still open? where can I buy X and what
tram do I get...? (piggybacking on wikipedia / dbpedia for types there
rather than re-inventing)
There are a million fun arty and gamey things to do too of course, and
I know the 'directory' stuff seems boring by contrast, but AR has the
potential to be an entry point to a world of information and right now
that information is differentiated by source rather than semantics. I
shouldn't have to hand pick a list of KML or whatever layers; machines
can be better at that stuff than us, but right now they're not getting
the help they need...
/braindump
cheers,
--Dan
Now you might say that those last scenarios are gen
Thanks to DanBri for the notes.
Phil.