Slashdot Mirror
Platinum Tech. Planning OSS Web 3D Tools?
There is a story over at News.com that says that Platinum Technologies has signed a
letter of intent with the Web3d Consortium to turn over the source code for some of the Cosmo tools for some type of open source developement project. This could be a big development for 3D on the web. What do you think of the Cosmo line or tools and programs? I tried them on an SGI (if it's the same code) a while ago and they were pretty sweet.
Here is the press release from Platinum..
The story doesn't mention that Platinum is doing this as part of their process of getting rid of all of their VRML assets and staff. They fired a thousand people on Tuesday, including everyone there working on VRML. *Then* they decided to talk about open-sourcing the Cosmo code.
A majority of the people who are/were heading the projet to create a next-generation VRML standard worked for Platinum and are now without employ. It remains to be seen whether this is the end of VRML development, or whether the VRML community can reorganize itself around open-source development.
I was reffering to PC's in general. VRML worked fine on my Cyrix 6x86MX 200, with a S3 ViRGE, and 32MB RAM, which was from '97, and Cyrix isn't known for it's great FPU perfromance. And that was in WINDOWS. Linux performance will be considerably better, especially if we can finally get widespread 3D acceleration. My FPU probably performed like a P100's... I'm willing to bet that most computers currently in use for what VRML might be used for (probably home use first, then possibly with video technologies a kind of teleconfrencing system for execs, all connected to the web) are in this catagory, or near enough to do the job. However, if you optimize the applications for K6-2/3/K7 or PIII's, then you'll definately get better performance.
Posted by CyberThrash:
3D is cool
Linux is cool
I've been dreaming of a designing 3d windowmanager for a while, but haven't had the technical experience or time to work on such a project. Maybe a (optimized)vrml-based windowmanager could become possible??
do you have raw polygons/sec or pixels/sec fillrate numbers? just for the fun to compare with PC stuff. :)
But MetroX/Xig don't render in hardware (although they'll be releasing ones that do Real Soon Now). AFAIS, the only real 3D you can get under Linux now is (still!) Glide+Mesa...
For those interested in a VRML public source project I'd like to hear from you. I was at VRML99 and I'm in the process of forming a working group to help manage a public source effort.
Currently we are focusing on two areas:
What license to use for the source. This is an open issue and must be worked out with Platinum. We are trying to find a license that will be best for the VRML community and for Platinum.
How to manage this particular project. Currently looking hard at many different models. Some of the issues are: source code versioning, review and codebase control, bug tracking/submission, updating of VRML spec, et el. So many issues, and I have no desire to repeat work already done.
We could use help from those who have travelled this path before!
If you are interested in helping please contact me or join the discussion on www-vrml. A mailing list will be setup soon.
-Alan Hudson
giles_@ameritech.net
Well, I work in a large CAD company, so I'm not the typical audience. But it is useful for our customers to be able to export VRML so that they can easily show people models in progress, without the viewer needing to start up Unigraphics (or needing to have a license for it in the first place).
It seems to take forever to 'move' anywhere in one of those environments. Granted, it might be sweeet on a SGI machine, but most of us don't have access to one.
VRML on my old 180MMX (purchased an overclock processor) was horrible. Quake wasn't that bad. Hmmmm... Maybe modify the Quake engine to include some VRML stuff.
And to make it easy, create a little program that'll convert VRML stuff to something the engine understands. Could I write this? Heck no. I'm an average coder who does something else besides coding for a living. Still, I don't see why it can't be done.
I haven't used Cosmo player on Unix, but I've used it on Windows and it's the best VRML viewer on that platform. I would be delighted if Cosmo were to become available for Linux and other unixes that aren't Irix -- open source seems to be the best way for that to happen.
As for the use of VRML on the web, the strength of VRML is that, cleverly used, it can pack a big media experience in a small file. For instance, with a 3k VRML file I can let you can fly around Star-trek style near the 100 closest stars to the Sun. That can fill up your whole screen and it's smaller than most people's stupid animated .gifs and the stupid drop shadow logos that people like to make with the GIMP. The trouble with VRML is that it's very hard to tell stories in 3-d. It takes a whole skill set in 3-d modelling, and anybody who is good at it could be making a lot more money making television commercials than they could make with the web. The web can only make a small amount of money off content, so web content must be ~very~ cheap to produce for a site to be profitable.
3D in general is still finding its feet in the computer world. there are all sorts of UI issues, and the hardware is just now getting to the point where mass deployment is feasible. But the potential is there--3D shared interactive games, for example. There was a company, NewFire, that was doing a very fast VRML engine that was good enough for interactive games.
Aside from that, VRML can be used as a file format for swapping 3D files around with others. The history of 3D is one of one-off tools optimized for one particular environment by one particular company.
I used to work at Intervista which subsequently became Platinum just as Cosmo did. Platinum layed off a large number of employees on Tuesday, including former members of the Intervista and Cosmo group. They both were almost entirely gutted with only a few developers kept on to integrate what they have with other Platinum products. 3-D in the enterprise, just what everyone needs, right? Hopefully VRML won't die an untimely death as a result of Platinum's acquisition of all major VRML tool producers and then subsequently disbanding them.
3-D on the web obviously is something whose time hasn't quite arrived or else one or both of these companies would have been able to survive on their own two feet. I'm sure there's a lot of good code to be had in the source when it gets released - they were a good, smart bunch of people.
Metro-X and XiG have had 3D servers out for some time now. I must say, I really like using XiG's AcceleratedX 3D.
"In true sound..." -Agents of Good Root
OSS VRML Tools, GLX, Precision Insight and HW accelerated 3D X Servers, XFree86 4.0 - It is all coming together - soon we shall have that 3D Virtual Reality distributed world system thingy so beloved of early-90s film makers (Johnny mnemonic,lawnmower man + all the rest)
Yippee!!! down with WIMP, I say!!!!
There's already a well-established standard for a collaborative networked 3D environment. It's got excellent image quality, and generally provides a virtual reality experience orders of magnitude better than any of the VRML implementations out there. The latency is well managed, and plenty of people are running servers for it.
It's called Quake.
Somehow, I think you are thinking of java and its ilk when you talk about dinky little 3d animations on a web page.
I wrote a weblication to convert molecular structure files to VRML-1 in order to be able to display them in a platform independent manner. Unfortunately the sorry state of VRML on Linux isnt helping the platform independence, but that is for another thread.
You can see some of the molecules I converted off of my page (click on the molecules button), or more directly off of the quickie weblication page for those with frames/graphics impaired browsers. I used either my tool (genVRML) or the wonderful VMD program which runs great on the SGIs and reasonably on the Linux machines.
I have also seen virtual museum walkthroughs on the web which allow visitors to get a feel for how they want to plan their visits. I have seen virtual astronomy and anatomy labs.
VRML is actually quite a good technology. Unfortunately you need a rather powerful machine with excellent graphics to do a good job at visualizing anything that is moderately complex. Your machine needs a fast OpenGL implementation and hardware texturing. This means once again, an SGI box or a nice Linux box with the accelerated X servers and the hardware stuff with Mesa.
I have yet to see *any* practical uses for VRML
in the web. Usually it is just some dinky little
3d object which one can freely rotate using the
mouse. It is usually nothing more than eye candy.
I still bear a grudge against Platinum for having to use their Apriori problem tracking system.
You'd think for half a million $, you could get an interface better than Pine.
--
Jason Eric Pierce
I actually told SGI reps 2+ years ago to go talk to Id...
In fact, Id should hire those developers and build on the Cosmo pieces. They have the tech and brains to fix VRML.
There ARE cool and useful things you can do with all of it.
sdw
Stephen D. Williams
VRML is derivative subset of OpenInventor file format. OpenInventor is a truly excellent interactive 3d c++ development library built on top of OGL.
SGI thought that vrml & the web would save their hides because everyone would start using 3D on the web, and that would somehow generate demand for SGI hardware. This failed miserably since nobody had any real reason to use 3D on the web, and nobody had any serious development tools to make it possible. If SGI had released OpenInventor source code back then VRML would really have taken off since developers would have been able to create excellent tools for building 3D worlds.
Instead SGI dropped OpenInventor (now licensed to www.tgs.com) and tried to create the end user modeling tools/viewers themselves. The Cosmo modeler was kinda crappy and expensive so nobody bothered.
The release of cosmo stuff is definite step in the right direction, but the real masterstroke that would revitalise VRML would be to release OpenInventor under LPGL. This would enable lots of developers to build proper 3D applications (not just web crap) and really would stimulate demand for good 3d hardware (ummmmmmm... SGI's VPC + decent drivers for Linux).
http://rareformnewmedia.com/
Ah well. Just an idea. Wonder what that was? (I was 12 at the time.)
--jon. Postel is dead. May we all mourn his, and our, loss.
Nah there's plently of dinky little 3d VRML objects around. You get a pencil or something and you can rotate it with your mouse. Great fun for about 2 minutes.
Funny, most of the VRML I have seen has been quite specific applications. Most of the Java I have seen has been stupid animated icons.
There is a nice jumping place to other VRML worlds, and it is reachable from here. The vast majority of worlds out there are not little play toys, which is not what you can say about the vast majority of java applets.
Sorry, but VRML in general, and COSMO in particular, suck.
Lots of academics producing a functionally poor, verbose, over complex load of rubbish.
To see it done properly:
http://www.viscape.com
And yes, it's proprietary and not available for Linux, but it's simply a 100% better way of handling 3D realtime environments. Also, it has some fantastic authoring software.
-----
yep, get one of these new TNT-2 or Permedia3 3D hardware accelerator cards and it flies.
software 3D is just a pain in the ass, no matter how fast CPU you got.
As a former co-founder and principal engineer at Newfire, I am compelled to comment on this...
:-), so we had to put in special case code. Doing this essentially restricts the VRML spec in places, so that these custom pieces of code can work.
Yes, we were able to implement BSP and other game-like technologies in our browser and authoring tool. But with VRML, it took a very long time to develop these enhancements.
From the technical point of view, VRML is too complex and too rich. It is a system built on a rich vocabulary and a few flexible way of combining the pieces (namely, the hierarchical scene graph + routing).
There are two ways to implement this system: one is completely genereal implementation very close in spirit to the specificaiton (i.e. pieces of code connected by routing). We did this at first, and I belive Cosmo Player and WorldView work like this.
However, this resulted in unacceptable performance (to us
And, once you've done this, one might ask: why use VRML in the first place, if, to make it run fast / look good, it's necessary to restrict its generality? Wouldn't it be easier to write a system that is more closely tied to the playback and authoring capabilities?
Another problem with VRML is it's lack of useful functionality. For example, there is still no way (AFAIK) to read mouse (only keyboard), no object-to-object collision detection (we had to implement another extension), no standard network protocolos, etc....
Perhaps at some point, CPU's and 3D HW will be fast enough to allow the "general" VRML implementation to work. But by that time, custom-coded systems like games and VR systems will look much better and offer more real world functionality that VRML will still be useless by comparison.
Kostya Vasilyev -- kostya@pobox.com
I use 'em on my SGI all the time - VERY sweet.
Could we use this type of technology to build a new 3D ultra-user friendly desktop for Linux?
What things would hold us back? I honestly don't have much experience in 3d world but I've used Blender before and found it to be quite fun.
A 3D interface could really be interesting. Any thoughts on the API or design?
Nick
Linux Systems Group
New York
That used to be a fairly popular SGI user interface demo a while back... you'd navigate through a hierarchy of 'buttons' that spin through 3D space, sort of like a flashy 3d program manager.
Don't remember any 3d window managers, though...
; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
This is a perfect timing-after the VRML consortium became Web3D, it's started to get noticed again. The Cosmo tools are simply the best that there are, PC's are finally powerful enough to do stuff like this, and Open Source can't seem to stop gaining steam.
Tim
What about scrapping the whole VRML thing, and creating a new markup-language? Just take the standards of vrml and start from there.
Doesn't VROML sound nicer? (O = Open).
--- "If a man speaks in a forest, and no woman hears him, is he still wrong?"
I would love to think this news will give VRML a big boost - and hopefuly a linux port - although it does feel a little bit late and after the fact.
;)
An often overlooked factor in the downfall of vrml is precisely *who* creates the stuff (the content) that actually makes people dl the plugin in turn creating demand for more and better VRML spaces.
The type of person who is able to make high quality, visually stimulating VRML content space would be a technically competent (programmer or near programmer - to understand the weird routing and syntax *and* be able to write javascript/java to implement anything more complex than the most soul sappingly dull interactions) with a good theoretical understanding of 3d preferably with experience in a "professional" package (ie. practical experience) - eg. Maya, Explore, Symbolics, SoftImage, Nichimen, Houdini, um, 3dStudio (...and playing with a crack you borrowed off your dodgy mate doesn't count!) who also has a firm grasp of the internet and ideally be comitted to open source (even pre-platinum announcement there was a profound scarcity of people willing to pay ppl high wages to create stunning work subsequently to be "given away" on a website - if you can view a VRML scene it's trivial to cut'n'paste objects others may have spent days even weeks honing down into a lightweight, good looking piece - companies think of this as "piracy" - rightly or wrongly) - oh yeah, so some lo-poly/games experience would also be good. Unfortunately those exact type of people are in such demand right now that they are currently being offered stupendous money to work in any of the big animation houses around the globe - *how many* people does it take to animate a tie-fighter, George? Disillusioned games animators would also provide a good source of suitable producers - unfortunately making games is too damn fun and VRML is too damn DULL - it would appear. Ditto film, ditto commercials...
Most professional animators fall **way** below the technical ceiling required to create bearable VRML. Animators are, by and large, "package" users - not the gourard-shading-by-the-seat-of-your-pants type of programmer/animator of five to ten years ago - programmer and artist brains rarely seem to arrive in the same head - most people who work in 3d animation tend to be either programmers or artists. VRML demands are far greater synergy between programming and artistry (not that programming isn't artistic, far from it - it's just less accessible)
The majority of VRML developers instead had great programming vision and practically no colour vision! (...or understanding of how to add detail without weight - or knowledge of the plethora of routine tricks, hacks, bodges and smudges that 3d animators use daily). Instead the VRML community was besieged with worthy projects of great technical interest and little visual merit (critically) and military stuff (those craaaazy army guys love anything VR!) and in between these a few genuinely worthwhile examples sneaked by (which encouraged me to keep following the www-vrml list).
Cosmo/SGI then shot themselves in the foot with such force it took VRMLs kneecaps by dropping development of CosmoWorlds on the SGI (anyone know whether SGI CW is OSS? I see *no* mention of it - AFAIK it still costs US$2300 a seat) ignoring the fact that 75% of professional animators use SGIs at work and find WindowsNT (the NT stands for NeanderThal) unpleasant and confusing. It's *far* easier to convince your boss to fork out a couple of grand on some software that lets your clients review your designs in 3d in *their* office, not yours - they *love* that kind of thing - than it is to either fork out a grand for some software stuck on your PC at home or run a moral/conscience probing crack which you can't use for production either. Oh, and CW sucks big time on the PC - I *can't* believe they left out the beautiful per-vertex colour editor !! fools !!
Meanwhile I'm off down the shops to pick up a big steaming Visual PC, dl my free copy of CW and start making some more of those free models for ppl to rip
I work for a company that - on the software side - used to produce a vrml authoring tool. In the 2yrs I've been here, I've seen the market go from somewhat active to almost dead. We stopped development on the product and don't push it much at all. Seems to me like vrml is in bad straights. Hmmmmm.... maybe opening the source will give it a shot in the arm.
On the other hand, I've yet to see any really practical application for the technology via the www. From what I've experienced, though this may have changed recently, vrml is dog slow.
"shop smart:shop s-mart" ash