Augmented Virtual Reality
Tuesday, February 20th, 2007 at 9:20 AM by: babbagelindenAt a very interesting session at Euro FOO last year on using Second Life in real life, Claus Dahl commented that Second Life is potentially a great platform for prototyping networked augmented reality applications. Everything in Second Life has an Id, everyone has a HUD and scripts in Second Life can use llSensor to scan the local area to read the Ids and llHTTPRequest to retrieve information from the web associated with those Ids. Second Life is a world in which everything has an embedded RFID tag and everyone can have an RFID reader, internet connection and augmented reality display for free. We can start prototyping very large scale augmented reality applications in Second Life today.
I’ve been meaning to experiment with these ideas ever since Euro FOOCamp, so it seemed appropriate to finally start hacking some LSL about when I needed a presentation for BarCampLondon2 last weekend.
Scanning the local area with llSensor and extracting Ids with llDetectedKey was trivial, but the party piece of many augmented reality displays is overlaying real world objects with virtual objects or information, so on Thursday night I set to work building a HUD which would overlay objects in Second Life with information in the HUD. In order to do this, the HUD script needs to do the same 3D projection that the viewer does thousands of times to render the world: transform objects from world space in to camera space and then apply perspective distortion to transform them in to 2D screen space. The standard way to do this is with a single matrix transformation, but because LSL lacks matrix support it has to be done manually: subtracting the camera position from the object position; rotating the object with the inverse of the camera rotation and then dividing the final X and Y co-ordinates by the Z value to perform the perspective distortion. A fly in the ointment was the lack of an LSL function to get the current field of view for the perspective divide, so I had to hard code it to 60 degrees. Luckily the centre HUD co-ordinate system extents are -1 to 1 regardless of the screen shape, so I didn’t need to worry about the screen aspect ratio.
After a few hours of hacking on Thursday night I had a HUD which scanned the local area for objects and moved to overlay their positions, but I still needed to retrieve information associated with each object and display it in the HUD. With only a day to go I needed a demo application that I could hack together really quickly, so I chose Digg. By adding a XyText display to the HUD to show the current rating of the augmented object, a pair of HUD buttons to increment and decrement the rating of the currently augmented object and llHTTPRequest calls to get and store the ratings I got it finished on Friday night and amazingly it all held together in a live demo at BarCamp on Saturday afternoon.
The major limitations of the demo version are that ratings aren’t shared between residents and that the button HUD stores absolute values rather than applying deltas, which would cause problems with simultaneous ratings by multiple people. Luckily Simon Willison came along to the “Show me how to build digg in <your favourite web platform here>” session that I proposed on Sunday and built a new back end in Django in 25 minutes, which will hopefully be live soon.
In the meantime I’ve released the demo version under the GPL which you can pick up from my spot in Ambleside to hack around with. I think there is huge potential for augmented virtual reality displays in Second Life. Gavin Starks from Global Cool commented at Euro FOO that one of the problems with carbon emissions is that no one can see them. By building a HUD and a database mapping SL object Ids to RL carbon costs we could make carbon costs visible in SL and use it to educate people about carbon emissions in RL. The Dublin and Amsterdam builds in SL also provide lots of potential for experimentation: pulling information about the RL locations from the web and displaying it on augmented reality displays in SL. I think we can have lots of fun with this.
UPDATE: Fixed a bug that caused incorrect positioning of the overlay information, so if you’ve been having problems, get a new HUD (thanks Mark!)
UPDATE: Ratings are now shared and viewable online at http://slateit.org, so please get a new HUD and start SLating!
UPDATE: Video demo of SLateIt here http://www.youtube.com/watch?v=dsfZpYXrVvM
February 20th, 2007 at 9:25 AM
Can we have a summary of what this means in plain English?
February 20th, 2007 at 9:35 AM
I’ve always been interested in the idea of using something like Google Earth, and some of those major city maps, being used as a foundation template that could be plugged into SL, so creative types could build a fictional ‘future’ city on top of at least SOME semblance of reality (San Francisco in the year 2100).
A good number of redevelopment agency types and real estate folks tend to take a slight interest in SL to augment reality, virtual tours, etc. Heck, even home improvement has a nice SL-RL crossover tie in.
February 20th, 2007 at 9:44 AM
What does this posting mean? It sounds interesting but I really have no idea what it means.
February 20th, 2007 at 9:56 AM
If you could meta tag objects, then you could take it a step further and dynamically generate in world models of real world things. The various “holodecks” are a good example, where a bunch of dynamically rezzing objects reproduce common “rooms” instead of just rendering a matrix of 2d pixels (aka an “image”)
February 20th, 2007 at 10:06 AM
I second Darth Juniper. I’ve read that long article and didn’t understand a single word of it. But it sounded interesting. If anyone could summarise it briefly in an intelligible language it would be nice :-).
February 20th, 2007 at 10:16 AM
It all sounds very exciting, but i have no idea what he was talking about. Me read book now!
February 20th, 2007 at 10:22 AM
If I get this right, we can display RL objects as seen on websites (photos?), together with whatever attributes we like to add (he has carbon emissions as an example - it could be an evilish green smoke around high emission objects such as streets with high traffic).
February 20th, 2007 at 10:30 AM
agreed. i too would appreciate an overview that skips the techincal aspects that i just won’t understand.
February 20th, 2007 at 10:30 AM
Good to see the discussion live on, Babbage!
A less dense review of what we talked about is this: There’s a trend in doing networked objects in the physical world. People are dreaming of constantly communicating things that you can actively exchange data with in the real world. Kinda like having a Star Trek tricorder.
That’s the augmented reality bit. What we talked about is how this is hard to do in the real world, because the physical bits are hard/impossible to get right. In Second Life that’s not a problem - so your Star Trek tricorder really works. You can actually measure everything precisely with perfect fidelity.
What Babbage has built is a tricorder for Diggs.
February 20th, 2007 at 10:32 AM
muito legallllllllll
February 20th, 2007 at 10:33 AM
eu sou muito legal simpatico extrovertido
February 20th, 2007 at 10:34 AM
Well, I figure this would be an interesting thing to see. Like visiting real-life landmarks in Second Life.
However, I don’t think we should implement this until the server overload issues are dealt with - this might simply make things worse for most people at the moment…
February 20th, 2007 at 10:39 AM
Shouldn’t you be working on mono. ;P
February 20th, 2007 at 10:44 AM
Umm, i guess a quick description here is that babbage is trying to create a HUD where objects or avatars could be “tagged” in some way, so information could pop up on your HUD describing an object postioned about on the screen where the object itself is. Something like hovertext, except that it is a HUD that is displaying the information only to you.
This would allow you to have a device that looks up additional information about an object, in this case the rating that other users of this demo HUD assigned a particular object.
February 20th, 2007 at 10:47 AM
I think what it means is to use RL models overlaying the world of SL to build things based on it meaning that it’s going to be a wonderful thing for those who want to recreate RL objects into SL.
February 20th, 2007 at 10:59 AM
You can think about the technology the same way as Google Toolbar tells you the rank of the pages you are visiting without actually the page to be aware of the rank it has. The page is “augmented” with additional information that comes from a third party source. In the case of the examples here, instead of pages, objects are augmented with rating information delivered by a system located in RL. The difficulty in understanding the idea was maybe because babbage started with a more abstract view of how this mechanism can be extended for other types of augmenting, not only object rating.
February 20th, 2007 at 11:10 AM
LOL @ Pagan
I’m sure this is one of babages ‘playtime’ projects and this is the place to tell us about it (after all, contrary to the whingers beliefs - who’ll no doubt be along shortly to rant and wail aboout grid performace - A) lindens are allowed spare time and B) this isn’t actually a technical issues announcement board, its THEIR blog to do with what they like).
Here’s a thought for the whingers though:
What does LL need most to succeed….money
Who can provide ‘insta-cash’……new investors
whats likely to make a whole new breed of investor stop, turn their heads and get out their wallets……an aspect of SL with powerful RL applications….such as maybe a demonstration that half the infrastructure is already in place for experimentation with augmented reality?
*sits back with a satisfied kitty grin* ;-p
February 20th, 2007 at 11:18 AM
Damn! I’d thought about coming along to BarCamp last weekend but couldn’t think of something nice to present …
gonna go play now ;-P
February 20th, 2007 at 11:37 AM
How about posting the source on the web somewere so us TGers cn have a play too
February 20th, 2007 at 11:47 AM
Note that llSensor & llSensorRepeat are grossly inefficient due to the mountain of Detected* data thats returned in bulk. Not only do these functions tax the simulators, the limitation to 16 results seriously hampers their potential usefulness, and forces some folks to use *more* sensors, which just taxes simulators even more.
There’s a page on the wiki called ‘LSL Useful Function WishList’ that’s been created here: http://wiki.secondlife.com/wiki/LSL_Useful_Function_WishList
More specific LSL sensor-like calls that return avatar & key information are discussed on that very page, such as llGetAvatarKeysOnParcel(), that would not be so sim-taxing, and return more results.
I hope the LL devs working with LSL are taking a look at this page, and taking the suggestions there into consideration. A little love here would be appreciated
February 20th, 2007 at 11:53 AM
Something I have wanted for a long time is the ability to add tags to any object. That would help considerably with sifting through inventory.
What comes to mind with this posting: (wishlist)
* ability to have private vs public tags for your own things
* ability for others to tag your stuff
* use of machine tags to encode things such as Geo-enabled information
for details see:
http://www.flickr.com/groups/api/discuss/72157594497877875/
http://radar.oreilly.com/archives/2007/01/flickr_launches.html
February 20th, 2007 at 12:11 PM
Who gave this nerd an acid tab?
I kept hearing wooshing noises over my head as I read this
February 20th, 2007 at 12:34 PM
OK thanks Onionpencil that made it a lot clearer. I think part of the problem is that Babbage did not define the meaning of terms like “augmented reality” anywhere in the article. But now I understand.
February 20th, 2007 at 12:46 PM
Isnt this the concept behind W.Gibson’s ‘Virtual Light’?
Personally Id find text to be a distraction, it would be like trying to read a book while moving around. How about a ’symbol cloud’ at a fixed position in relation to the avatar or object, where each icon (from a restricted set) would encode information, perhaps clicking on them would bring up textual info? In a multi lingual environment text is too restrictive and exclusive but an icon isnt. The HUD could give options for the icon set to display thus cutting down on screen clutter.
Perhaps residents could be persuaded to be a bit more specific in their interest tab on the profiles, and fill out some structured metadata about themselves. The hud could then give everyone in your field of view an ‘aura’ or symbol cloud that is coded to select those individuals that share common attributes or have attributes that you select as interesting or to be avoided.
The virtual equivalent of wearing perfume or a digital pherenome.
February 20th, 2007 at 12:46 PM
Sorry Darth, I’ve added links to wikipedia for “augmented reality”,”RFID” and “3D projection” which should hopefully clear things up. I also encourage you to just pick up a pair of HUDs from Ambleside, put them on and have a play.
February 20th, 2007 at 12:49 PM
I’d love to pick up a HUD but I can’t, being on the Teen Grid and all. Any chance of the scripts being made avaliable elsewhere?
February 20th, 2007 at 1:07 PM
Can you put this on the TG.
February 20th, 2007 at 1:12 PM
Where exactly on Ambleside?
February 20th, 2007 at 1:23 PM
Huh?
February 20th, 2007 at 2:56 PM
It’s the box with the “green thumb” when you arrive at the SURL
February 20th, 2007 at 3:53 PM
You might also find James Cascio’s presentation called Participatory Panopticon interesting too.
http://www.itconversations.com/shows/detail375.html
He describes what a world might be like when we have HUDs in real life.
February 20th, 2007 at 4:31 PM
GET THE RESTRICTION MEASURES AT WORK. The lag is killing again.
And please set them at 25.000 cause thats the number you Linden can really handle. Everything above gives (occasional) problems.
February 20th, 2007 at 6:15 PM
Eu nao consigo jogar….quando vou logar fala q o capslook esta ligado mais realmente nao está alguem me ajuda
February 20th, 2007 at 9:32 PM
Using The Meshverse To Address Global Warming
Just a few days ago, was talking about Augmented Reality With Croquet, now from the Linden blog I see a post about Augmented Reality that talks about how Second Life can be used to deal with global warming issues:
Gavin Starks from Global Cool commente…
February 21st, 2007 at 6:13 AM
Sorry Babbage
I got and tried the HUD. I have the red and green “thumbs up” buttons at the top, permananently.
Whilst on your land I had a flat green “thimbs up” icon with a 0 above it, which did seem to move about a little, but no sign of any size change as it moved, let alone any “3d rotation”. Just a simple icon wandering about, but not pointing at anything particular that i could see.
I teleport home, to a quiet island sim, and the floating icon is now gone. I detach/reattach. Still nothing. Not even when I “view transparent”.
When it first worked on Babbages patch it didn’t seem to do anything which bore much realation to the descriprtion. Now it doesn’t work at all.
Please babbage, spend 5 minutes posting a simple description of what we should actually see happen with your HUD. Forget the generaised augmented reality philosophy for the moment. What should we see ? How should it work ? Why did I only see it whilst on your land ? Which object is it selecting ? Tip of thumb ? Can we choose (or even know) which object it is selecting ?
I do appreciate this is no more than a crude “proof of concept” prototype, but we cant eveluate it if we don’t know even how it is supposed to work, and therefore judge if it is working correctly.
February 21st, 2007 at 8:15 AM
I’ll try plainer English here… I have studied a little on augmented reality systems. I think it is easier to explain if you picture it outside of Second Life. Augmented Reality is made possible by wearing something like a virtual reality system. The difference between Virtual Reality and Augmented Reality is that with Virtual Reality you are cut off from Reality. Augmented Reality overlays the two, so you would have information or data displayed while still being aware of most, or all of reality. In a real-life system this information is likely uploaded and downloaded from your viewer via tiny micro radio transmitters, embedded in products, vehicles, people, or whatever..
This post talks about simulating such a system within Second Life. Such a system is possibly closer to realization, in some form or another, then you might imagine. Some companies already include these chips in products, to track inventory, customers locations, behaviours, etc. The darker side is a global ID system based on such chips. Imagine this: autonomous A.I. robot soldiers (such as those currently being developed by DARPA and Boston Dynamics under The Bush Administrations generous funding), could maneuver around an environment scattered with these RFID chips. The world we see would appear to them as a map based on these points, for instance they could search and differentiate targets based on this data: hunting down the unchipped “terrorists” and leaving be the chipped “citizens”. Hopefully a never used concept, but just one way this tech can be used. These RFID chips are already in our cellphones and passports. Hopefully they have a brighter picture then the idea I envision here and hopefully my explanation is a little clearer
February 21st, 2007 at 8:56 AM
This sounds academically interesting and very clever. I wish I could code like this. Am I right in thinking that it would be a bit like using infra red goggles at night to Augment your vision, but you could get more info, like say, in some games when a vehicle approaches you you get a display that tells you its current position, course, speed and perhaps type? But, presumably this would have a plethora of other as-yet-unthought-of applications.
I mean like extending our senses, to be able to smell as efficiently as a dog, or “see” in 3D, like dolphins use ultrasound? Or, maybe have a small lens which shows close up detail in the centre of the viewer, like some birds.
Or is it more like using your pocket PC to surf the web while you walk around, like someone once told me a certain famous coffee shop was thinking of sending (or spamming) cappucino coupons via bluetooth if you walked near their store.
In fact, is it virtual bluetooth?
I have more questions than answers, maybe if I could see it? Could you trap the video and put it on youtube or somewhere?
I don’t know what it is, but I think I want one anyway… but it sounds interesting, and very clever, so have 10 smug points.
Does this mean I can have RL anklettes and armbands and do real dancing into SL?
Teledildonics?
RL HUD’s in my reading glasses, so I can stay in SL all the time?
February 21st, 2007 at 9:23 AM
Just for reference for the non-technical folk, I thought I’d dive into the applications (very) briefly…
Augmented Reality centers around just that–enhancing your current perception of the world around you. The classic implementation of this would be in a sort of visor worn over the eyes that would overlay additional information served up via a computer worn somewhere else, likely the back. (Few of these concept machines ever made it off the ground though, and this will likely be the case until Pattern Recognition software and computer hardware become a lot more sophisticated.) As for the information displayed, it varies. A lot. For example:
- A person who’s allergic to peanuts walks into a restaurant. With a quick adjustment to his AR system, all the menu items that would trigger his allergy are highlighted in red, or possibly the ’safe’ dishes in green.
- A traffic cop is driving down the road, his windshield reporting the speed of every car within his field of vision by displaying the number where it appears to be over their car.
- You’re wandering the aisles at a shopping mall, and your visor displays small symbols on or near every item you see, telling you how other consumers rate the product, as well as the price of said item and the location of the best competitor’s price.
- As you stop in a small Asian food market and pick up a can of Lychees (a type of fruit), you press a couple buttons on the side of the visor and it immediately presents you with a short list of recipes using them, limited to those that you’re missing no more than a couple other ingredients for, and also limited to the ones that match your tastes.
The applications are *very* endless. Unfortunately, these would depend on large-scale distribution of RFID chips and readers, as well as some very nice computer hardware, and a wireless data network (like city-wide Wi-Fi maybe?) to support all the info gathering. That’s why this tinkering is going on… Second Life provides a convenient environment to try out these concepts in. Imagine:
- Everyone who ventures onto your parcel is flagged red if their profile has no payment data attached.
- People are flagged yellow if they’re currently carrying a usable weapon, defined as any running scripted attachment with push capability.
Lots of interesting concepts for SL, too. I’ll definitely be checking this out.
February 21st, 2007 at 12:12 PM
suzie sue, the hand icon should move around, but not scale or rotate. It shows ratings of 0 for all objects it augments initially, but if you click the green or red hand buttons and wait 30 seconds for the data to reload you should see the ratings change. If you teleport to another sim or parcel the HUD should start augmenting objects again after 30 seconds, unless the parcel doesn’t support scripts or you’re more than 30m from any objects.
February 21st, 2007 at 2:44 PM
Sensors are so well implemented that they hardly lag at all and are very usefull (lies)
It’s a real pain dealing with avs using more CPU then the rest of a sim. Some way to meter usage would be nice instead of letting lag bombs walk around with 20 scanners.
Carry on.
February 21st, 2007 at 7:25 PM
I found the article interesting and the picture of me in the Blarney stone very exciting!!!Thank you Babbagelinden:)
February 22nd, 2007 at 1:52 AM
Babbage, I have read this line of blog with true satisfaction, as we are involved exactly in this way of doing augmented reality mixing virtual worlds and real world for years.
Two years ago we patented at the EPO some devices named “3DFORM-ID” and “3D-LABELS”, now commercially named “3DTurtles” that makes exactly that: conecting directly the reality to a virtual world for doing virtual replicas of our environment in real time, for doing millions (yes, millions) of applications concerning artificial intelligence, automatic driving, true efficient robotics, virtual true tourism, security, er…all!…
A 3D Turtle is basically a device with a huge memory and with 3 or more spatial references, that contains the 3D shape description of the real object to be downloaded, by example by a robot, for being used in the creation of a virtual replica of the environment.
Including not only an ID number but the 3D shape into the real object is a giant advantage, as allows total independence and safety to the individual object, then prolonging the use of the object´s ID forever, on despite network conection will not be available or inclusivelly if the object´s maker is out of business.
If you are interested pls visit our workbench site at http://www.3dturtles.com or at our web site http://www.deespona.com
then enter in the 3DFORM-ID chapter in the main page.
February 22nd, 2007 at 8:00 AM
Digg Virtual Objects From Within SL?
Well, I was all ready to anoint Babbage as my favorite Linden, but I can’t get his gadget to work. This post on the official Linden blog describes a HUD that lets you Digg objects inside the virtual world of Second Life, if I’m understandin…
February 22nd, 2007 at 8:03 AM
My head hurts, I just got used to VR let alone AR
February 22nd, 2007 at 8:40 AM
*calls Opus Flamand*
February 22nd, 2007 at 2:38 PM
Well, Babbage, I for one know exactly what you are talking about.
I recall seeing prototype for these types of devices at SIGGRAPH in the 1980s. This would be akin to wearing goggles that projects a graphical information that overlays the scene viewed by the eye. Imagine an automobile windshield that highlights curb and centerline positions in bad weather. Consider the combat HUDs in fighter aircraft. Recall the snazzy robot vision presentations in the Terminator science fiction movies.
In secondlife this means having a HUD device that moves parts of itself around the display so that little mobile gadgets appear overlaid on the VR image. All the talk of camera positions, matrix algegra and projections is simply saying that the gadgets have to stay near their associated avatar even as the view changes. All the sensor and database stuff is simply finding the avatars and their positions and then dredging up a little information about them to be displayed.
Congratulations. It is a remarkable tour-de-force and highlights the incredible flexibility of the Linden Scripting Language.
des/demona
February 23rd, 2007 at 3:09 AM
yes, mixing real world objects with virtual worlds is an old topic. The key question is how to connect the real world object position, orientation and shape with your virtual scene, and a lot of erratic and unstable work have been done around, as RFID numbers related to centralized 3D datasets, videocamera edge recognition proceedings that only works with very simple objects, spreaded 3D pinpoints as in the mocaps (motion capture systems) etc.
A 3D turtle ( http://www.3dturtle.com )is our solution to that, as this devices gives a fast and bullet proof solution: the own object gives you, forever and in any moment and environment, their shape, their position and orientation for being included in a virtual scene created by a detection system. The IP number is used as an additional guarantee for checking info in a centralized database in the network, to see if the detected object is really with this shape. A system based only in a IP number referenced to a centralized database have HUGE problems concerning legal issues, not only about privacy but about civil responsability, security, etc. By example, a centralized IP server can´t works for fully automatic driving in the real world, as an eventual failure in the network (electrical or data conection, makers dissapearance from the market) can be risky to an unacceptable level with an additional dispersion of civil responsability. The same can be said about robotics, etc. And for this reason the pharmaceutical industry, by example, is implementing more and more info in bigger memories in their RFIDs identified products (usage, composition, legal terms), as networked ID numbers are so legally light in many questions. ID numbers are ok for verification, and for playing in a virtual game, or working in a close environment as a supermarket chain, but if a generic reality is concerned (real cars, real robots, real people that can be everywhere in any environment) then we have the necessity to include also all the info with the own real object, or we will have true legal problems for doing real world applications based on this mixing with virtual worlds.
February 24th, 2007 at 4:41 PM
[...] Jim Purbrick talking about SecondLife being used as a prototype for real-life innovation [...]
April 14th, 2007 at 11:56 AM
K…aside from all this…WHEN am I gonna get a tier payment bill?
I’ve changed my billing three times with PayPal and have done the same with my SL account…STILL no notice of billing from SL to PayPal.
I can’t believe I’m going through this a SECOND time…March was the first time I’ve EVER had trouble with payments in the almost three years I’ve been at SL!
LL has ALWAYS gotten their money…for land purchases, for tier payments and for membership fees.
I can’t believe that my ‘Second Life’, the place I count on to go to for relaxation is doing this!
I sure hope someone ELSE is going to do a ‘game’ like this soon…the trampling of little feet LL is gonna hear will be the hordes of rat/avies jumping ship!
All that will be left will be piles of rat/avie turds!
July 14th, 2007 at 3:26 AM
[...] A Second Life article that caught my attention: “At a very interesting session at Euro FOO last year on using Second Life in real life, Claus Dahl commented that Second Life is potentially a great platform for prototyping networked augmented reality applications. Everything in Second Life has an Id, everyone has a HUD and scripts in Second Life can use llSensor to scan the local area to read the Ids and llHTTPRequest to retrieve information from the web associated with those Ids. Second Life is a world in which everything has an embedded RFID tag and everyone can have an RFID reader, internet connection and augmented reality display for free. We can start prototyping very large scale augmented reality applications in Second Life today. Augmented reality (AR) is a field of computer research which deals with the combination of real world and computer generated data. At present, most AR research is concerned with the use of live video imagery which is digitally processed and “augmented” by the addition of computer generated graphics. Advanced research includes the use of motion tracking data, fiducial marker recognition using machine vision, and the construction of controlled environments containing any number of sensors and actuators. Posted on Tuesday, February 20th, 2007. Link: Augmented Reality / Second Life Blog [...]
September 25th, 2007 at 1:15 PM
augmented reality? well this is the definitive confirmation of my SL and Google earth convergence theory and probably the bigger SL competitor in a near future
I hope the following step will be my 3DTurtles concept (www.3dturtles.com) massive implementation.
This is on the news, folks:
“Rumors of Google’s plans to create a virtual world that rivals that of Second Life have popped up once again over the weekend. The company could now be collaborating with Arizona State University to test the 3D social network, which may be tied into Google’s current applications of Google Earth and Google Maps.”:
http://games.slashdot.org/games/07/09/25/1437249.shtml
http://arstechnica.com/news.ars/post/20070924-google-testing-my-world-for-launch-later-this-year.html
March 12th, 2008 at 12:19 PM
Cool post.
For more groundbreaking augmented reality demos see my latest post: “Top 10 augmented reality demos that will revolutionize video games”
October 15th, 2008 at 9:35 PM
i want very mach play secont life. i’m from grecce. good bye\\\…:}