Commercial Interest In Open-Source 3D Environment

cellulama writes "Is virtual reality back? A commercial vendor has started developing for Croquet, which is an open-source tool for collaborating and sharing data, with an emphasis on 3D visualization. The system was developed at the University of Wisconsin-Madison for research and "co-creativity," but a company named 3Dsolve is looking for military applications. What's next -- open-source America's Army?"

Comments

Open-Sourcing the Army

[Justice8096]

I have yet to see any animosity towards open source in the military. They like open source because you don't have to wait 6+ months to get the purchase completed for proprietary stuff, and you don't have to worry about license problems when dealing with vendors ( the licenses can become a hostage towards keeping the contract ).

Re:Open-Sourcing the Army

[Anonymous Coward]

Nothing like an army of bearded men to instill fear and terror!

just kidding, we love you alan cox and rms.

On a more serious note, is there any other open source bearded gods?

Re:Open-Sourcing the Army

[ProfKyne]

You must not be familiar with the Navy's NMCI then. It was basically an initiative by industry lobbyists to get the Navy to standardize on Windows 2000 for everything, along with a number of other commercial applications.

Re:Open-Sourcing the Army

[Justice8096]

Mandates work fine, as long as they are in the budget. Anything done outside of the budget isn't allowed to spend money. :-)

Re:Open-Sourcing the Army

[ProfKyne]

No, we were told that we couldn't use anything other than Windows 2000 (no Mac, Linux, etc). Not even on our development boxes. Even though Linux is free.

Possibility of Open Source 3D CAD?

[pipingguy]

Go for it.

Croquet

[Scott7477]

On the web page this is described as a completely new OS. What would prevent a new open source OS from gaining widespread acceptance now that running multiple OS's on one machine is common? The web page has set forth some ambitious goals for this project. If it lives up to the goals it should take off.

Re:Croquet

[mwielgosz]

What would prevent a new open source OS from gaining widespread acceptance now that running multiple OS's on one machine is common?

There many Open Source OSs that dont gain widespread acceptance.
Linux, is still one of them. Although is is more widely accepted than OpenBSD

;-)

Re:Croquet

[pkhuong]

Yeah, but this one could just run as a user process. So if the need is there, it could gain acceptance much more easily. [and then they could subtly add in more and more applications until people are ready to just switch to Croquet. Mwahahah]

West Point a croquet partnet

[pkhuong]

http://croquetproject.org/Community/consortium.htm l:
has a list of Criquet's core education partners, which includes (in the middle of institutions more usually associated with CS)... United States Military Academy at West Point. I wonder what _they_ see in it. We know the military have always been interested in 3D and VR, but why do they see a need to for croquet rather than a normal (and probably better suied to 1st person view and simulation) 3D toolkit? Dynamism and good introspection aren't really needed to simulate airplanes, etc.?

On second thought... Agents fit particularly well in smalltalk's message passing paradigm... So does croquet's goal of seamless networking (iirc?), especially when simulating hundreds or thousands of agents. Smalltalk, the new lead (as in Pb) figurine! :)

It won't work

[dshaw858]

I worked on Croquet with colleagues at the San Diego Supercomputer Center... it's some pretty gnarly code, written mainly in Smalltalk. I don't think that a commercial vendor would be able to deal with it; people who love code, yes... people who just want money from it... no.

- dshaw

Re:It won't work

[Anonymous Coward]

Do you like other smalltalk programs? Ie., is it gnarly because it's poorly designed, or just because it's smalltalk?

And are the APIs gnarly, or does the gnarliness just come out if you're actually developing Croquet, not just using it?

(Not trying to be a wiseass, I'm genuinely interested in your take on it. I've played around with Squeak a bit, considering much more but haven't taken the plunge yet.)

croquet is the next generation smalltalk

[Anonymous Coward]

I believe if you want to get a sense
of croquet, it would be best to look at
a previous creation of some of the same people,
squeak (an implementation of smalltalk-80)!

Re:West Point a croquet partnet

[smeat]

Please contact R. Craig Hogan for your writing reeducation.


smeat!

Interesting

[Refrozen]

I have been 'keeping up' on Virtual Reality, and I see it as a very close thing... Especially after reading the Tom Clancy: Net Force books (great books BTW).

What really gets me, is that a google for virtual reality turns up this as the first result. (it's neat, but seems kinda unrelated)

OSS is neutral on use...

[Spoing]

OK, more specifically free software is neutral on how the software is used and who can use the software. Open source software tends to follow that as well.

This means that if a mailing list manager is created by an ardent advocate of choice...the mailing list licence does not have a clause that says 'can not be used by right to lifers'. Or bisa-versa. If it does, it's not OSS or free software.

Drug pushers, dictators, mass murders, bunny skinners, and traffic violators are treated the same as military or private users -- no matter what you think of any of these groups.

Madison is too cool

[togofspookware]

> The system was developed at the University of Wisconsin-Madison

All the coolest stuff happens in my hometown, but instead of going to school there I shipped off to Platteville, where the hicks and the imaginary engineers reign. Dumb me :P

Distributed events... worrying

[CrosbieFitch]

Distributing events is too fragile.

They're still over-fixated on synchronisation.

Gotta distribute state man. It's the only way.

Re:Distributed events... worrying

[CrosbieFitch]

The short version is that we have to ultimately have a distributed atomic transaction on our objects. In other words, at some point the system of replica objects has to commit to the same state everywhere in the same pseudo-time and as close to possible in the same real-time. If things go wrong -- and there are many kinds of things that can go wrong in a highly distributed system of machines and networks of different speeds and reliability -- we have to make sure that when the dust clears either the entire new version got computed, committed to and labeled, or no part of that transaction happened at all (just like in a banking transaction, we can't have money being moved from one account to another be disturbed by any kind of error or crash. The source has to decrement and the destination has to increment, or neither)

NOoooooooooooo!!!

You're not dealing with bank accounts. You're dealing with reality - it only has to appear to be consistent.

The key idea in TeaTime is that the state of objects evolves through a distributed two-phase commit protocol. Behaviors of all objects that influence each other are first computed, contingent on completion of all dependent object behaviors, and then those behaviors are atomically committed. If the behaviors are not completed in time, all contingent calculations are undone by the individual objects.

There simply isn't enough time to ensure consistency across multiple nodes. Or rather, it's a combinatorial explosion problem and thus non-scalable. And don't tell me you can partition the world into neat little isolated chunks - reality is continuous.

Each node simply has to do the best it can, and live with the fact that it will not produce quite the same result as its neighbours. Hence why you need to distribute state and not events.

One contractor comes to mind...

[sethstorm]

If one does some research on General Dynamics, you'll see that they have done 3d tank simulations. If this appears to be anything related, and if the licensing is right, I'd bet on them picking it up if they can figure a way to keep it secret.

Surprise !

[Wudbaer]

I don't think that a commercial vendor would be able to deal with it; people who love code, yes... people who just want money from it... no.

Surprise ! There are people in commercial entities that love code, too ! *GASP* And they even sometimes can make money from this ! *BOOM**Head explodes*

we could also OUTSOURCE the army!

[museumpeace]

Demo's have to be done with least cost so the Army often sees alpha versions of stuff that vendors hope to sell running on RedHat. But when the money comes down, its for Solaris or maybe Windows NT or 2000...open source still scares the brass in the pentagon. After open source, the other trend in software [but one that is a worry for programmers rather than proprietors] is outsourcing...now that IS something I would like to see our DOD pursue more vigorously.

Re:we could also OUTSOURCE the army!

[Anonymous Coward]

Hmm...you know, there are good strategic reasons not to outsource your military software...

Re:we could also OUTSOURCE the army!

[museumpeace]

I was thinking more of infantry jobs than the software. I actually am a dot.com refugee. I holed up in a defense industry SW job because I know my clearance lets me do work they are never going to sub out to Bangalore Binary Ltd.