W3C home > Mailing lists > Public > public-ar@w3.org > September 2011

Re: [AR Standards Discussion] ARML Standards Working Group beingformed through OGC (available for comment)

From: Rob Manson <roBman@mob-labs.com>
Date: Fri, 16 Sep 2011 13:42:14 +1000
To: Carl Reed <creed@opengeospatial.org>
Cc: Neil Trevett <ntrevett@nvidia.com>, cperey@perey.com, public-declarative3d@w3.org, discussion@arstandards.org, W3C AR Community Group <public-ar@w3.org>
Message-ID: <1316144534.21759.270650.camel@robslapu>
Well that's what I call a productive thread 8)

Carl...as for your point 4 below.

>         4. WRT Web RTC, I do not see a conflict. Again if the ARML
>         extension capability is defined properly, an ARML payload
>         could easily reference a Web RTC session or encoding.

I don't think that's strictly true.  The RTC relationship with AR is not
really just about sessions and encoding.  It's more about access to the
stream data and optimising for processing like the MediaStream
Processing API draft that Jens shared on twitter yesterday.
http://hg.mozilla.org/users/rocallahan_mozilla.com/specs/raw-file/tip/StreamProcessing/StreamProcessing.html 

So just plugging an RTC session/stream into ARML is not going to cut it.


Neil, on a slightly related note I'd love to hear more about Khronos'
plans for and progress with the typed array spec too.
http://www.khronos.org/registry/typedarray/specs/latest/
>From browsing the web it seems to be gaining acceptance with w3c related
specs...but it's not easy to tell if this is really the case?


roBman



On Thu, 2011-09-15 at 20:47 -0600, Carl Reed wrote:
> Neil -
>  
> Thanks so much for the quick and incisive feedback. 
>  
> I agree that we need to work on terminology to avoid confusion. We are
> using sensor fusion as defined in the DoD/Intel and environmental
> domains, so a different focus for sure.
>  
> Based on your feedback, I agree that the activities are synergistic.
> If I understand correctly, we could easily "wrap" the StreamInput API
> with OGC SWE interfaces and payloads and then integrate into existing
> or planned SWE based applications. This would be really cool and
> definitely would provide added value throughout the workflow.
>  
> As to OGC Fusion work, here is the link to the follow-on report for
> Phase II of the sensor fusion project:
> http://portal.opengeospatial.org/files/?artifact_id=41573
>  
> Thanks again. I have the information I need to brief the OGC SWE
> community and hope we can collaborate on these efforts so that there
> is seamless flow of content throughout the stack.
>  
> Regards
> 
> Carl
>  
>         ----- Original Message ----- 
>         From: Neil Trevett 
>         To: Carl Reed ; cperey@perey.com ; roBman@mob-labs.com 
>         Cc: public-declarative3d@w3.org ; discussion@arstandards.org ;
>         W3C AR Community Group 
>         Sent: Thursday, September 15, 2011 6:42 PM
>         Subject: RE: [AR Standards Discussion] ARML Standards Working
>         Group beingformed through OGC (available for comment)
>         
>         
>         Hi Carl,
>         
>          
>         
>         StreamInput
>         
>          
>         
>         >>  Well, now I can ask why is Khronos developing yet another
>         sensor API when the OGC (and other groups) have had a mature
>         set of sensor standards that are already widely implemented on
>         a global basis?
>         
>          
>         
>         Great question, and I think it has a very straightforward
>         answer J
>         
>          
>         
>         Khronos standards focus on the low-level interface between
>         hardware and software and as you say StreamInput is 100%
>         focused on onboard device sensor interaction.
>         
>          
>         
>         We did look at the available literature such as:
>         
>         http://portal.opengeospatial.org/files/?artifact_id=36177  
>         
>         and we are all using the term senor fusion in rather different
>         ways – we might want to clear up that terminology.
>         
>          
>         
>         Anyway, StreamInput sensor fusion is concerned with things
>         like abstracting the gyro, accelerometer and compass under an
>         API to enable sensor vendors to innovate delivering a holistic
>         positional data stream that is faster and more accurate than
>         looking at each sensor separately.  We have proven proprietary
>         APIs from some of the industry-leading sensor hardware vendors
>         being contributed as input to this exercise.
>         
>          
>         
>         This seems to be a very different level in the software stack
>         than SWE or other sensor network standards where fusion refers
>         to combining images and features from different sensor
>         sources.
>         
>          
>         
>         I think the activities are synergistic.  Using StreamInput
>         hopefully will make it much easier for innovative application
>         to generate appropriate data for SWE in a device portable way.
>         
>          
>         
>         I definitely agree we should coordinate to ensure those
>         potential synergies are realized – that’s exactly why Khronos
>         is participating on the AR Standards Group.
>         
>          
>         
>         If you are aware of other industry initiatives working at the
>         level of StreamInput please let us know, we will be glad to
>         collaborate/outreach.  We definitely don’t have enough spare
>         bandwidth to re-invent any wheels or re-solve problems.
>         
>          
>         
>         Declarative 3D
>         
>          
>         
>         >> I read the charter for the W3C Declarative 3D for the Web
>         Architecture. I really do not see much conflict here in terms
>         of the OGC ARML work and the W3C community activity
>         
>          
>         
>         I agree.  This group is working to use existing browser
>         machinery – such as DOM – to bring programming 3D in the
>         browser to a higher and more familiar level than raw WebGL for
>         a typical web developer.  It is still a general purpose way to
>         program 3D – and will likely drive down into WebGL for
>         acceleration without a plug-in.
>         
>          
>         
>          
>         
>         Best regards,
>         
>          
>         
>         Neil
>         
>          
>         
>         Neil Trevett
>         
>         Vice President Mobile Content, NVIDIA | President, Khronos
>         Group
>         
>         2701 San Tomas Expressway, Santa Clara, CA 95050 USA
>         
>         M: +1 (408) 464-7053 | O: +1 (408) 566-6512 | F: +1 (408)
>         986-8315
>         
>         ntrevett@nvidia.com | www.nvidia.com
>         
>          
>         
>         From: discussion-bounces@arstandards.org
>         [mailto:discussion-bounces@arstandards.org] On Behalf Of Carl
>         Reed
>         Sent: Thursday, September 15, 2011 12:32 PM
>         To: cperey@perey.com; roBman@mob-labs.com
>         Cc: public-declarative3d@w3.org; discussion@arstandards.org;
>         W3C AR Community Group
>         Subject: Re: [AR Standards Discussion] ARML Standards Working
>         Group being formed through OGC (available for comment)
>         
>         
>          
>         
>         Well, now I can ask why is Kronos developing yet another
>         sensor API when the OGC (and other groups) have had a mature
>         set of sensor standards that are already widely implemented on
>         a global basis? Implementations by Ericsson shows how OGC SWE
>         standards work with the onboard sensors in a Smartphone. There
>         is also the ISO/IEC JTC1 Working Group 7 Sensor Network
>         activity that ISO and the OGC have been involved in for a
>         couple of years. 
>         
>         
>          
>         
>         
>         Anyway, the Kronos activity appears to be focused on
>         Smartphone onboard sensors. Would be nice to engage in some
>         collaboration with Kronos on this activity so that we do not
>         have conflicting but instead complementary standards.
>         
>         
>          
>         
>         
>         I should add that OGC Members have been working various
>         aspects for sensor fusion for quite some time. Non-trivial.
>         There are a number of publicly available reports.
>         
>         
>          
>         
>         
>         As to the OGC ARML activity, a couple of observations:
>         
>         
>          
>         
>         
>         1. I see the ARML and W3C PoI activity as complementary as
>         long as they both use the same information model. The KML
>         point geometry model is based on ISO 19107 and GML so there is
>         consistency there as well as with the IETF location object
>         geodetic encoding (which will be used for the Next Generation
>         911 system in the US) and the OASIS point geometry encodings
>         for a number of OASIS standards. We definitely do not want to
>         have a new point geometry encoding model!
>         
>         
>          
>         
>         
>         2. I read the charter for the W3C Declarative 3D for the Web
>         Architecture. I really do not see much conflict here in terms
>         of the OGC ARML work and the W3C community activity.
>         Complementary in many ways actually, especially in the content
>         sharing and content creation topics. Further, the OGC and the
>         Web3D Consortium have a Memorandum of Understanding to enable
>         closer collaboration. Further, Fraunhofer is also an OGC
>         Member.
>         
>         
>          
>         
>         
>         3. WRT KML, when Google first submitted KML into the OGC
>         standards process, there was no extension mechanism. This is
>         one of the elements that was added into the standard. My point
>         here is that if properly designed and expressed, an ARML
>         extension mechanism would allow for incorporation of of
>         reference to other AR standards and payloads. I worked the
>         EDXL standards activity in the OASIS Emergency Management TC.
>         EDXL is an emergency management exchange language. Much time
>         was spent on their extension mechanism. The extension
>         mechanism is flexible enough to either incorporate or
>         reference GML, KML, satellite images, or other geospatial
>         payloads. Very nice.
>         
>         
>          
>         
>         
>         4. WRT Web RTC, I do not see a conflict. Again if the ARML
>         extension capability is defined properly, an ARML payload
>         could easily reference a Web RTC session or encoding.
>         
>         
>          
>         
>         
>         Just some thoughts. 
>         
>         
>          
>         
>         
>         Regards
>         
>         
>         
>         Carl
>         
>         
>          
>         
>         
>                 ----- Original Message ----- 
>                 
>                 
>                 From: Christine Perey 
>                 
>                 
>                 To: roBman@mob-labs.com 
>                 
>                 
>                 Cc: public-declarative3d@w3.org ;
>                 discussion@arstandards.org ; W3C AR Community Group 
>                 
>                 
>                 Sent: Thursday, September 15, 2011 8:45 AM
>                 
>                 
>                 Subject: Re: [AR Standards Discussion] ARML Standards
>                 Working Group being formed through OGC (available for
>                 comment)
>                 
>                 
>                  
>                 
>                 
>                 Hi,
>                 
>                 This integrative work is precisely one of the purposes
>                 of the AR standards community. 
>                 
>                 But, it only works when/to the extent that people want
>                 it to. 
>                 
>                 Gentle reminder that the next meeting is Oct 24-25 in
>                 Basel. 
>                 
>                 The OGC ARML activity will be topic of a presentation
>                 and discussion, but the other groups which Rob
>                 mentions (W3C DAP, W3C Web RTC, W3C Audio WG) are not
>                 on the agenda...
>                 
>                 
>                 
>                 Christine
>                  
>                 Spime Wrangler
>                  
>                 cperey@perey.com
>                 mobile +41 79 436 6869
>                 VoIP +1 (617) 848-8159
>                 Skype Christine_Perey
>                 
>                 
>                 On 9/15/11 2:58 PM, Rob Manson wrote: 
>                 
>                 I think ya knygar raises a very interesting point for you Martin and the
>                 OGC too.  How do you see this relating to all the work already under way
>                 for web based AR standards development.
>                  
>                 I mean how would this integrate with the Declarative 3D work?
>                 Or the POI WG work?
>                 Or the DAP and Web RTC work?
>                 Or the Audio WG work?
>                  
>                 And how would this integrate or leverage the StreamInput work that
>                 Khronos are starting?
>                  
>                 I'm all for standards...but I think before we head into another set of
>                 weeds I'd really like to see our overall community doing more
>                 integrative work.
>                  
>                 roBman
>                  
>                  
>                 On Thu, 2011-09-15 at 12:29 +0000, ya knygar wrote:
>                         Hello Martin Lechner!
>                                 I strongly disagree that AR standards are still not required.
>                         i don't see any soul here - with ignorance for IT standards,
>                         i think what Blair MacIntyre - the developer of another useful AR
>                         standard - exactly mean:
>                          
>                                 Given that a vast amount of what would be "in" an ARML or KARML data stream, there is absolutely no chance any of them will be compatible with each other any time soon, so why not work on the big issues before going down into the weeds?...
>                                 to work together to be compatible where we agree, and go our own way when we don't, and then see things evolve basic on real people actually
>                         doing things with the various browser and so on.
>                          
>                          
>                         -
>                                 I think that - while the "Web Story" is a little bit different from the "AR Story" - it still makes a good "reference story".
>                         1. Following your context -- do you envision some AR Net rather than
>                         functioning only in the standards defined - Web?
>                         (given the currently strong approach on standardization of "Device
>                         API's", i mean - at least 3 serious groups - working for the 'next'
>                         Web)
>                          
>                         2. Do you think it is 'Ok' to make some other consortium and move
>                         separately from the current W3C governance?
>                         (like WHATWG did, for example)
>                          
>                         3. Could you, please, elaborate on the differences where are the good
>                         old, decentralized
>                         "World Wide Web (WWW, or simply Web)" as "an information space in
>                         which the items of interest, referred to as resources, are identified
>                         by global identifiers called Uniform Resource Identifiers (URI)."
>                         model does not fit / where it fits in your opinion?
>                          
>                                 Once again, I'd like to invite everyone (as in *EVERYONE*) to work within the ARML 2.0 SWG
>                         do you plan an open mailing list or forum during the development of proposition,
>                         so people - who aren't the members of OGC for one or another reason
>                         - would be able to contribute/etc. into the standard formation - in other way?
>                          
>                         sincerely,
>                         knygar
>                          
>                          
>                         2011/9/15 Martin Lechner <martin.lechner@wikitude.com>:
>                                 Hi Blair, Carl, Rob et al.!
>                                  
>                                 While I do agree that AR is not used by masses of people yet, I strongly
>                                 disagree that AR standards are still not required. In my opinion, a standard
>                                 the AR community agrees on will help the industry grow significantly, if (as
>                                 in *IF*) the standard takes into account that it will require extension in
>                                 the future. Still, we all know that AR applications are out for quite some
>                                 time now (with a lot more to come every week), and I guess all of us will
>                                 agree that they all have significant overlaps in their functionalities. As
>                                 far as I'm concerned, this already justifies working on a standard for AR.
>                                 Figure how HTML was created - it started out with a couple of tags, and I'm
>                                 pretty sure Tim did not know precisely how the Web will be shaped in the
>                                 future. Yet, it was extensible, and turned out to be successful. I think
>                                 that - while the "Web Story" is a little bit different from the "AR Story" -
>                                 it still makes a good "reference story".
>                                  
>                                 In my opinion, it's about getting things started, allowing the AR industry
>                                 to agree on a standard, while still not closing doors for extending the
>                                 standard. It will be one of the key topics in the ARML 2.0 SWG where we need
>                                 to ensure that future AR requirements can be met (by adding new components
>                                 to the standard), I keep thinking about a component model where various
>                                 components can connect with the existing ones.
>                                  
>                                 Once again, I'd like to invite everyone (as in *EVERYONE*) to work within
>                                 the ARML 2.0 SWG to define an AR standard within the OGC. You guys at GA
>                                 Tech could certainly contribute a lot to the success of the SWG, so in case
>                                 you are still interested, we will kick-off the ARML 2.0 SWG in the OGC TC
>                                 meeting in Boulder on Monday, Sept. 19th. Whoever wants to join and get
>                                 involved in the SWG is invited!
>                                  
>                                 Best,
>                                 Martin
>                                  
>                                  
>                                 Am 04.09.2011 15:26, schrieb Blair MacIntyre:
>                                         Hi Martin,
>                                          
>                                         I agree with Rob;  if you have a larger list of efforts, it would have
>                                         been more useful to include them, rather than making it appear quite so
>                                         "wikitude-centric".  Folks will be far more interested in contributing if it
>                                         appears to be more inclusive;  as it stands, the document feels a bit to
>                                         focused on your company, which won't serve you well.  Witness my reaction.
>                                         ;)
>                                          
>                                         We'll be happy to discuss the directions we are going to be going this
>                                         year with KARML;  the current implemented version touches on some of what
>                                         you are going after, and our plans for Argon for this year touch on much of
>                                         the rest of it.
>                                          
>                                         Georgia Tech is not a member of OGC as far as I can tell, so our
>                                         involvement won't be "formal".
>                                          
>                                         Just so you know, I feel that this effort is premature;  I find it ironic
>                                         that you are taking KML (a "standard" that evolved from a widely used
>                                         defector standard into something more formal only after it was proven to be
>                                         useful), and using it as the basis for a "design before we really know what
>                                         people will use" standard.   I use "we" inclusively:  I don't think any of
>                                         us (including researchers like me) really _know_ what needs to be in these
>                                         standards and tools, since AR is still not being used by very many people
>                                         for very many things, and certainly not in the architectural scenario these
>                                         standards will impact.   There are some things that can be standardized,
>                                         perhaps (e.g., some of the ARML 1.0 things, which we've taken further in
>                                         KARML, like extending ideas of location reference beyond LLA).  But when you
>                                         start talking about "events" I get nervous.
>                                          
>                                         I'd much rather see an informal effort by those of us (you at wikitude, my
>                                         team, perhaps others) who are actually building on top of KML and building
>                                         javascript libraries for AR, to work together to be compatible where we
>                                         agree, and go our own way when we don't, and then see things evolve basic on
>                                         real people actually doing things with the various browser and so on.  Given
>                                         that a vast amount of what would be "in" an ARML or KARML data stream, there
>                                         is absolutely no chance any of them will be compatible with each other any
>                                         time soon, so why not work on the big issues before going down into the
>                                         weeds?
>                                          
>                                         On Sep 4, 2011, at 8:24 AM, Martin Lechner wrote:
>                                          
>                                                 Hi Rob, hi Blair!
>                                                  
>                                                 We already have a list of other standards/efforts we will include in the
>                                                 charter prior to the startup of the SWG, and KARML is on the list already,
>                                                 along with others. The revised list will be published in an updated charter
>                                                 document after the public comment-period. I agree that KARML is valuable
>                                                 contribution towards an AR standard.
>                                                  
>                                                 As a general "Call for Participation", I would love to have
>                                                 representatives from other institutions which proposed AR standards in the
>                                                 SWG, it would be great to have you on board. However, as far as I
>                                                 understood, you need to be OGC member to work within an SWG, this is a
>                                                 formal requirement.
>                                                  
>                                                 In case you consider joining OGC to work within the SWG, highly
>                                                 appreciated - I think Carl is the one to talk to about it.
>                                                  
>                                                 Regards,
>                                                 Martin
>                                                  
>                                                 Martin Lechner
>                                                 CTO
>                                                 Wikitude GmbH.
>                                                 +43 (0)676 840 856 300
>                                                 martin.lechner@wikitude.com
>                                                  
>                                                 You are catching me underway ... On my iPhone!
>                                                  
>                                                  
>                                                 On 04.09.2011, at 14:09, Rob Manson<roBman@mob-labs.com>  wrote:
>                                                  
>                                                         I think those are fair questions that hopefully Martin or even Carl,
>                                                         Steven or any of the OGC people on the list here could address.
>                                                          
>                                                          
>                                                         roBman
>                                                          
>                                                          
>                                                         On Sun, 2011-09-04 at 07:57 -0400, Blair MacIntyre wrote:
>                                                                 Interesting.  How do we comment on it if we aren't OGC members?
>                                                                  
>                                                                 Obviously, the complete lack of any mention of our work on KARML is a
>                                                                 bit surprising (if only in the "other know efforts" section), considering
>                                                                 it's more mature than either ARML or ARchitect, is well documented on our
>                                                                 website, and has a fully working implementation in the iTunes app store
>                                                                 (Argon).  And, of course, since I know they know about Argon and KARML, it's
>                                                                 clearly an intentional omission.
>                                                                  
>                                                                 While I realize their bias is toward their own commercial interests, it
>                                                                 would seem to undermine the position of OGC as a standards organization to
>                                                                 have a small group of people leverage them as a platform to promote their
>                                                                 commercial product.
>                                                                  
>                                                                  
>                                                                 On Sep 4, 2011, at 4:07 AM, Rob Manson wrote:
>                                                                  
>                                                                         Here's a publicly accessible link.
>                                                                          
>                                                                              https://portal.opengeospatial.org/files/?artifact_id=45439
>                                                                          
>                                                                          
>                                                                         Thanks Carl/Steven.
>                                                                          
>                                                                          
>                                                                         roBman
>                                                                          
>                                                                          
>                                                                         On Thu, 2011-09-01 at 00:27 +1000, Rob Manson wrote:
>                                                                                 Hi,
>                                                                                  
>                                                                                 cross posting this from the AR-UX list as I think many will find it
>                                                                                 interesting/relevant.
>                                                                                  
>                                                                                      Augmented Reality Markup Language (ARML) Standards Working Group
>                                                                                      being formed. Draft charter available for review/comment if
>                                                                                      you're an Open Geospatial Consortium (OGC) member.
>                                                                                  
>                                                                                      Please address any comments or questions to Martin Lechner -
>                                                                                      martin.lechner@wikitude.com This is the start of a 3 week review
>                                                                                      period. After this period, Carl Reed [OGC CTO] will do a formal
>                                                                                      call for participation. Also, if your organisation wishes to be
>                                                                                      represented as a Charter member of this new Standards Working
>                                                                                      Group (SWG), please let Martin and Carl know.
>                                                                                  
>                                                                                      I realise a number of you are not and may never be members of
>                                                                                      the OGC, so this is just some market information for you. Any
>                                                                                      resulting standards from the OGC are freely available.
>                                                                                  
>                                                                                  
>                                                                                 http://www.linkedin.com/news?viewArticle=&articleID=730135900&gid=3844396&type=member&item=67968411&articleURL=https%3A%2F%2Fportal%2Eopengeospatial%2Eorg%2Ffiles%2F%3Fartifact_id%3D45285%26version%3D1&urlhash=1ywF&goback=%2Egde_3844396_member_67968411
>                                                                                  
>                                                                                      /via Steven Ramage @ OGC
>                                                                                  
>                                                                                  
>                                                                                 roBman
>                                                                                  
>                                                                                  
>                                                                                 _______________________________________________
>                                                                                 Discussion mailing list
>                                                                                 Discussion@arstandards.org
>                                                                                 http://arstandards.org/mailman/listinfo/discussion
>                                                                                  
>                                                                         _______________________________________________
>                                                                         Discussion mailing list
>                                                                         Discussion@arstandards.org
>                                                                         http://arstandards.org/mailman/listinfo/discussion
>                                                         _______________________________________________
>                                                         Discussion mailing list
>                                                         Discussion@arstandards.org
>                                                         http://arstandards.org/mailman/listinfo/discussion
>                                                 _______________________________________________
>                                                 Discussion mailing list
>                                                 Discussion@arstandards.org
>                                                 http://arstandards.org/mailman/listinfo/discussion
>                                 --
>                                 - - -
>                                 Martin Lechner
>                                 CTO
>                                  
>                                 Wikitude GmbH
>                                 Ginzkeyplatz 11
>                                 5020 Salzburg/Austria
>                                 Phone +43 662 243310
>                                 Mobile +43 676 840 856 300
>                                  
>                                 http://www.wikitude.com
>                                  
>                                  
>                                 _______________________________________________
>                                 Discussion mailing list
>                                 Discussion@arstandards.org
>                                 http://arstandards.org/mailman/listinfo/discussion
>                                  
>                                  
>                         _______________________________________________
>                         Discussion mailing list
>                         Discussion@arstandards.org
>                         http://arstandards.org/mailman/listinfo/discussion
>                 _______________________________________________
>                 Discussion mailing list
>                 Discussion@arstandards.org
>                 http://arstandards.org/mailman/listinfo/discussion
>                  
>                                            
>                 ______________________________________________________
>                 
>                 _______________________________________________
>                 Discussion mailing list
>                 Discussion@arstandards.org
>                 http://arstandards.org/mailman/listinfo/discussion
>                 
>         
>         ______________________________________________________________
>         This email message is for the sole use of the intended
>         recipient(s) and may contain confidential information.  Any
>         unauthorized review, use, disclosure or distribution is
>         prohibited.  If you are not the intended recipient, please
>         contact the sender by reply email and destroy all copies of
>         the original message. 
>         
>         ______________________________________________________________
Received on Friday, 16 September 2011 03:42:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 16 September 2011 03:42:47 GMT