[openARweb] W3C Workshop: AR on the Web

[Phil Archer]

Hi everyone,

I joined this group as soon as I heard about it (through Dan Bri) but
haven't posted yet.

I'm on the W3C Team and am looking into what we might be able to do to
help make open AR a reality. Along with Dan Appelquist (Vodafone),
Rittwik Jana (AT&T) Christine Perey (independent), I'm chairing a
workshop in this area in June.

Full details of the CfP are at [1]. I tried to include a link to this
Google group in the CfP but for some reason our blogging software threw
a hissy fit and wouldn't let me.

The discussion about choice of CRS is very pertinent to the workshop.
More generally, what is missing from the Web standards stack that should
be there?

Signs are promising for a positive workshop so please take a look at the
CfP and, if you can submit a paper by the end of May that would be terrific.

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


--
Subscription settings: http://groups.google.com/group/openARweb/subscribe?hl=en

[ThomasWrobel]

Greetings, thanks for joining.
Its good to see all the major party's are slowly getting into the
discussion.
Its hearting to know everyone seems to not only be keen for open-AR in
principle, but are
actually willing to start to come together to help pin it down. I'm
really looking forward to developments.
I'm not sure if I'll be able to make it myself to Barcelona, but with
luck I should at least be able to put out a paper or even an demo of
our AR over WaveFederationProtocol concept.

Some random thoughts;

The interesting thing with AR, unlike the Web as it is now, is that it
isnt page-based. The idea of a "page" in real space is kinda
arbitrary, and I think getting our heads around that might take
awhile.
HTML has evolved to be less about pages too, of course. And features
such as websockets will make server<>client streaming a lot easier.
But still, as with AR we are never really going to be looking at
pages, but rather streaming most content based on location and line of
site, maybe its best not to go for html based data exchange at all.
(of course, thats not to say web pages shouldnt be, geo-tagged, but
for tieing 3d models to gps positions, a more "per-object" stream
might be better)

The other difference is with the viewing. Most of us these days use
tabbed-browsing for the web. We keep many pages open at once and flick
between them.
With AR though we might be viewing content from many sources at the
same time, each as a layer overlaid on or FOV. So its more
simultaneous viewing of various sources.

Cheers,
Thomas :)

On Apr 26, 3:21 pm, Phil Archer <p...@philarcher.org> wrote:
> Hi everyone,
>
> I joined this group as soon as I heard about it (through Dan Bri) but
> haven't posted yet.
>
> I'm on the W3C Team and am looking into what we might be able to do to
> help make open AR a reality. Along with Dan Appelquist (Vodafone),
> Rittwik Jana (AT&T) Christine Perey (independent), I'm chairing a
> workshop in this area in June.
>
> Full details of the CfP are at [1]. I tried to include a link to this
> Google group in the CfP but for some reason our blogging software threw
> a hissy fit and wouldn't let me.
>
> The discussion about choice of CRS is very pertinent to the workshop.
> More generally, what is missing from the Web standards stack that should
> be there?
>
> Signs are promising for a positive workshop so please take a look at the
> CfP and, if you can submit a paper by the end of May that would be terrific.
>
> Cheers
>
> Phil.
> --
>

> Phil Archerhttp://philarcher.org/
> @philarcher1
>
> i-sieve Technologies                   |      W3C
> Sentiment Analysis Beyond Impressions  |      Mobile Web Initiativehttp://i-sieve.com                    |      http://www.w3.org/Mobile

[Mike Liebhold]

Dear Phil,

This is great news that the W3C is opening up an exploration of AR. I
would lover to participate in your workshop, but am unlikely to be able
to afford the travel. So, FYI here are a few, among many, key
considerations, aside from concerns of CRS, for the W3C including:

* an AR web is a focal plane view of the linked web, so all views of all
digital objects will ideally be semantically aligned by the a W3C
framework for viewer markup. This is complicated by the fact that AR is
also a focal plane view of -map- data, aligned within OGC Frameworks (
open geospatial cosortium, ) So, one key task of a W3C work group is
explore how to convene a collaborative approach with the OGC to define
AR software stack, starting perhaps with the notion of an AR browser, or
viewer.

* One specific point of contact with the OGC, might be the definition of
standard 'Place Pages' . Currently google is racing ahead promulgating
their version of 'Place Pages' ( based on KML - an OGC product). Ideally
a place page would be a fully functional web page, following W3C
standards and practices, but so far google's version is a greatly
constrained document. I believe ( and hope) that google would cooperate
in refactoring the description of a standard 'Place Page'

* Although I am unconvinced, Several people on this list argue that HTTP
is inadequate for realtime AR, and that a version of the xmpp protocol
is better suited. I'm certain this asserting will be widely argued here
on this list. Still, there is are many open issues related to the future
of AR protocols. e.g. what are the variations of RESTful services for
AR? ( How would RESTful concepts apply to other protocols?) What
specifically are the requirements for linked data for AR services?

* . . . [ more to follow]

Cheers,

Mike Liebhold

On 4/26/10 6:21 AM, Phil Archer wrote:
> Hi everyone,
>
> I joined this group as soon as I heard about it (through Dan Bri) but
> haven't posted yet.
>
> I'm on the W3C Team and am looking into what we might be able to do to
> help make open AR a reality. Along with Dan Appelquist (Vodafone),
> Rittwik Jana (AT&T) Christine Perey (independent), I'm chairing a
> workshop in this area in June.
>
> Full details of the CfP are at [1]. I tried to include a link to this
> Google group in the CfP but for some reason our blogging software
> threw a hissy fit and wouldn't let me.
>
> The discussion about choice of CRS is very pertinent to the workshop.
> More generally, what is missing from the Web standards stack that
> should be there?
>
> Signs are promising for a positive workshop so please take a look at
> the CfP and, if you can submit a paper by the end of May that would be
> terrific.
>
> Cheers
>
>
> Phil.



--

Subscription settings: http://groups.google.com/group/openARweb/subscribe?hl=en

[ThomasWrobel]

As Mike suggests, some of us our a little concerned about the dive
towards HTTP as a base's for a AR Protocol.

I'll just lay down my thinking along these lines here for now, but it
might be worthy of a separate conversations thread later.
Part of this is just to get it clear in my own head too, so excuse for
some of the baby-steps in my explanation.
--
HTTP and HTML are, as everyone here knows ancromins for "HyperText
Transfer Protocol" and "HyperText Markup Language"
Letting use jump from page to page pretty seemlessly.

We have built on these standards with stuff like php, allowing on-
demand server-side generation of pages.
And Javascript, client-side manipulation of pages at the Dom level.

But its still all bound by the concept of a page, which is defined
explictly by what the creator of the page sets.
The size of the page, the quantity of content and who contributes to
that page is all determained by the person
that wrote the HTML.

However with Augmented Reality there is no boundarys to the screen.
The positioning of elements stretchs to infinity in all directions.
What you see is dertermained by your camera view, and your devices
draw-distance.
The idea of having a boundary in AR just doesnt exist. There is no
"edges".

So how would you dertmain what AR data to put on the HTML page? A
whole worlds worth on a single page? Have seperate pages for each
city? town? Street?
Would clients need to load two pages if they were on the edge between
two locations?

Now imagine the contents of this AR View.
This content, just like a HTML page, has to be downloaded.
So what is the most efficianct method of loading the objects for this
view? How can the servers pass the data to the client with the least
overhead?

These are the questions any standard should be seaking to answer. And
I cant help but think using HTML/HTTP is the easy answer, but not
necessary the correct one.

For instance, theres no obvious equilivents to hyperlinks in AR Space.
Instead its the clients location that determains whats in its field of
view, and thus what it needs to see.
HTML, on the other hand, provides a fixed bundle of data.

Sure, you could use javascript in HTML to then stream other AR data
in, but it would be *that* information that would be the AR protocol,
not the initial (2D) page set up.
And if your defining a streaming protocol based on the users location,
then you wouldn't need javascript at all...any client-side code would
do.
--

So, to some up; maybe I'm wrong in my belief that WFP specifically
would make a suppiour protocol for geolocateing data. I dont want to
state boldly we have the answer, as it may yet prove unworkable.

But I certainly think we shouldn't just assume that building on HTML5/
HTTP is automaticaly a good choice.
Its the principles of openness that are important to keep as the Web
evolves, but we shouldn't be scared to format our data differently
when the end experience itself will be so different.

Thanks for listening,

Thomas Wrobel :)

[Tish Shute]

Hi Phil,

Great to see this event in Barcelona. I also would not have a budget
to travel there but I will definitely consider submitting a paper.
Will the proceedings be recorded or streamed?

best,
Tish

[Christine Perey]

Hi Tish,

Thanks for your interest in the W3C Workshop on AR on the Web.

There is no plan for video streaming or capture at this time, however,
I've touched bases with co-chairs and we could do it if there is demand
and a way to fund it (or to do it at no cost). There was also discussion
about having second life support.

The dialog is usually "scribed" in real time and minutes of W3C
workshops are always published, and all the position papers will be
available to public as well.

--
Christine

Spime Wrangler

cpe...@perey.com
mobile +41 79 436 68 69
VoIP (from US) +1 (617) 848-8159
Skype (from anywhere) Christine_Perey

[Christine Perey]

Hi Thomas,

Thank you for articulating your concerns.

In my experience with the W3C Social web Incubator Group [1] over the
past year as well as my contact with the W3C over the year+ prior to
that, the purpose of a workshop (the one which I co-chaired in January
2009 is documented here [2]) is not to mandate what type of standards
are necessary or on what basis (past or other standards) they might be
developed.

The purpose of the workshop in June in Barcelona is to have discussions
precisely of the nature which you have so graciously begun below, and
others which are of interest to participants. The participation at the
workshop is expected to be varied and represent many different points of
view and interests (geolocation, visual recognition, hardware, software,
etc).

At the end, or after the conclusion of the workshop there will be
discussion among those who feel most strongly about if and how the W3C
might be involved in driving standards, but that is not the sole or
primary purpose of the workshop.

I hope that others will express their interest/concerns about or support
for using HTML for AR.

Regards,

--
Christine

Spime Wrangler

cpe...@perey.com
mobile +41 79 436 68 69
VoIP (from US) +1 (617) 848-8159
Skype (from anywhere) Christine_Perey

[1] http://www.w3.org/2005/Incubator/socialweb/
[2] http://www.w3.org/2008/09/msnws/

[Dan Brickley]

Hi all

Great to have this forum :) just a quick comment for now re W3C Workshops

On Thu, Apr 29, 2010 at 11:27 AM, Christine Perey <cpe...@perey.com> wrote:

Hi Thomas,

Thank you for articulating your concerns.

In my experience with the W3C Social web Incubator Group [1] over the past year as well as my contact with the W3C over the year+ prior to that, the purpose of a workshop (the one which I co-chaired in January 2009 is documented here [2]) is not to mandate what type of standards are necessary or on what basis (past or other standards) they might be developed.

The purpose of the workshop in June in Barcelona is to have discussions precisely of the nature which you have so graciously begun below, and others which are of interest to participants. The participation at the workshop is expected to be varied and represent many different points of view and interests (geolocation, visual recognition, hardware, software, etc).

Yes, W3C Workshops are really an exercise in 'fact finding' and community outreach, typically conducted as a precursor to any possible new work in unexplored areas. If you look at what the W3C Process Document says about them, http://www.w3.org/2004/02/Process-20040205/events.html you'll notice an emphasis on timing,

"Workshops and symposia generally last one to three days. If a workshop is being organized to address the pressing concerns of Members, the Team must issue the Call for Participation no later than six weeks prior to the workshop's scheduled start date. For other workshops and symposia, the Team must issue a Call for Participation no later than eight weeks prior to the meeting's scheduled start date. This helps ensure that speakers and authors have adequate time to prepare position papers and talks."

...this comes also from an attempt to be as inclusive as possible, while acknowledging that not everyone can afford to fly to Workshops they'd love to attend. W3C also has a pretty strong culture of accountability and documentation, so online minutes and results are generally looked after quite carefully.  And there's often a real-time IRC chat channel, although twitter/statusnet seems to be increasingly taking over as the medium for real time textual asides.

For me the biggest challenge facing AR standards is scoping; AR is the meeting place for several technology worlds, rather than a single self-contained area of work. The piece I am most interested in progressing is the shared problem of describing the real world in sufficient detail / precision that AR interfaces can exploit those descriptions. We had some great discussions on this around the Amsterdam AR Dev Camp, and I owe the world a little writeup of that! Will circulate it here asap...

cheers,

Dan 

[Phil Archer]

I'd like to add my thanks to Thomas and Mike for kicking off the discussion.

Christine and Dan have made several comments about the W3C process that
I hope adds clarity. An aspect of the discussion so far that surprises
me a little though is the implicit assumption that W3C is only about
HTTP and HTML.

I need to make two points about that:

1. As Dan says, workshops are essentially about fact finding, not about
setting out a pre-determined W3C position or a plan to get everyone to
use HTML5 for everything. Let me make this clear: W3C does not have a
position on AR!

The workshop is about seeing what the lay of the land is, seeing if
there is a community demand for work to be done that could usefully be
done under the W3C's banner. We don't actually initiate /anything/. All
we do is to respond to the community, particularly our members. There
are three possible outcomes from the workshop:

- a finding that there is no significant community demand for work
that W3C should/could be involved in;

- an incubator group. That is, a short-term, light weight group that
produces a report within no more than a year setting out what standards
work could be done. An incubator group (known as an XG) does not itself
create any standards although it can speed up the work of...

- a full working group chartered to produce a defined set of standards
and ancillary documents.

It is not for me to pre-judge what the actual outcome might be but my
suspicion is that an XG might be the best vehicle for taking the work
forward.


2. The idea that W3C is only about HTTP and HTML is demonstrably
incorrect. Almost as an aside, HTTP is defined by the IETF, not W3C
(although there is, of course, close liaison). HTML, whilst clearly a
flagship standard, is not the only one. Others include XML, widgets, Web
Apps, the linked data stack (Semantic Web), the Web Services stack,
video on the Web, CSS, SVG plus areas such as internationalisation and
accessibility. The bit that I work in, ubiquitous Web, is all about the
Web on everything other than desktop/laptop computers.

The choice of protocol - be it HTTP, XMPP or something else - is
orthogonal to the debate from a W3C perspective (honestly!).

In summary, I am very reluctant to express any view on where the
conversation might go. That's the point of the workshop - to find out.

So, again, thank you for kicking off the conversation. I will do my best
to track it here and elsewhere and hope very much that you will indeed
submit a paper to the workshop, even if you're unable to join us in
Barcelona.

HTH

Phil.