Slashdot Mirror
Engine for Collaborative Science Education MMOG?
plisdku asks: "My research advisor wants to create an online, collaborative science education game for middle- to high-school students; we need a 3D engine! A pioneer in internet-assisted laboratories (iLabs), he has developed real experiments that students can operate remotely over the internet. We envision a game where students wander a campus, solving Caltech-style 'stacks,' and learning collaboratively as they perform the great experiments of the history of science in simulated and real laboratories. What we need is a 3D engine and content creation system with which our team of artists and bright student developers can produce a prototype within a year. Requirements: customizable avatars, flexible integration with other technology (iLabs, CompEdu, etc.), and reasonable licensing fees (open source would be great). Our last project used SGI tools -- what are our modern-day options?"
Flash.
Best Windows Freeware
if I'm not mistaken, id has released the quake 1 and 2 engines under the gpl
Crystal Space
It's highly configurable and easy to code for. There are multiple included implementations of apps using the engine that come with it so you can keep the wheel reinvention to a minimum.
Also it supports multiple formats to be imported such as 3DS and the like.
MTW
The Quake 2 engine is probably the most advanced open source engine youl'l find. But It sounds like it would be better to start from scratch if your advisor intends to be putting science expiraments into this...
I've never used it, but Ogre3d looks promising.
http://www.ogre3d.org/
This would appear to mesh exactly with your aims :-
http://www.opencroquet.org/
Two psossible downsides SmallTalk and Alan Kay. Both are positives for me but I understand some peoples aversion.
It's free, but not open source:
http://www.qubesoft.com/
Second Life is ideal for this sort of thing.
They even have a special program for universities.
It's not just a graphics engine, it's a whole 3D world where you can collaboratively build and script stuff... it has very customizable humanoid avatars and is extremely user friendly.
It's got physics too - based on the "industry standard" Havok engine.
Chris Dede's group at Harvard is doing something like this:
http://muve.gse.harvard.edu/muvees2003/
I saw him present this at AERA this year.
You have to root around to find the science experiment stuff.
Open-source under the LGPL license, and developed at the Naval Postgradute School (former home of America's Army).
Please post on the forum if you have any questions, and I'll be glad to help you out.
Ranked #1 at DevMaster.net's 3D Engine Database, and with all of the features that you expect from the likes of UT3, HL2, etc all for $150. http://www.truevision3d.com
Torque engine from Garage Games
.NET based quickly points out runs on Mono.Net.
RealmForge from www.realmforge.com is open source. They just released v0.6.1 but are not yet to a beta version, (Puts on Tinfoil Suit) and it is
Very complete feature set (many yet to be implemented however).
Nevrax
Radan Labs
Reality Factory
3D Game Studio
but better, browse to
http://www.devmaster.net/engines/
and waste some time there!
*click**beep**beep* Scotty, One to Mod up!
The Cube engine is open source, I'm not sure how well it would fit your requirements, but its probably worth a look. It runs really well on lots of older hardware as well!!
http://cube.sourceforge.net/
.sigs are for losers
i would think that the engine that 3 rings open sourced would be a good fit for this type of project.
all java based and puzzles would be replaced by experiments.
and bigger install base if it does not requires a super hot video card.
depends on the content i suppose.
I've found Soya to be a very friendly Python-based 3D engine. It also has ODE physics built in, so that might be useful, too. It also works closely with Blender models. In addition, the developers are very responsive, and they've produced a number of tutorials to get people up to speed quickly. Not sure if the software you'd like to integrate with has Python bindings, but this is a good option if it does.
You may want to check out Nevrax.org home of Nel, the MMO Engine that powers The Saga of Ryzom. NeL is licensed under the GPL and is a set of C++ libraries and utilities.
She loves me: 09F911029D74E35BD84156C5635688C0 She loves me not: 09F911029D74E35BD84156C5635688BF
I'm a graduate of Caltech and I assure you that "Caltech" and "stacked" are two words that do not go together. There are a few cute girls there but none of them are stacked by any stretch of the imagination.
... is going to cut it from what I gather. Most engines, no matter what features they laud tend to do really well at certain functional aspects and only mediocre to dismal in others. Another thing to consider is that the engine as a whole will only be incorporating 3D visuals as the visualizations portion of the engine. I would gather that the networking portion of the engine would have to be very robust, efficient and all around top-notch if they're planning on doing a MMOG style persistent world.
This effects what your presentation portion needs to handle. Complex, smoothly animated avatars are probably NOT going to be something needed... especially when you consider large numbers of people congregating in any one are of the persistent world. Not only will you have to be keeping up with all that information, but you're going to have to render it too. What this means is that all the advanced graphics methods presented by engines similar to Doom3, HL2, or Unreal3 are probably overkill.
You also have to consider the types of machines that schools would purchase to access this world. Creating high demands on visualization and processing power in general means higher prices for not only the server and production machines, but the clients as well. And let me tell you, schools don't like paying big bucks for computers if they don't have to.
A Quake 2/3-like engine may suffice for visualization needs, but I guarantee you that the networking would still have to be loads better as neither engine scales well to extreme server populations (ie. 100+... they make the assumption that you need to keep track of all players at all times rather than only those players 'close-by'). Luckily the Q2 engine is GPL'd and so can be the subject of modification on that end... however there may be more suitable engines that I'm simply unaware of.
You might want to give the torque (tribes2) engine a try, as it supports large environments, unmodified up to about 128 simultaneous players, has decent physics and is more mature in a lot of ways than anything in the F/OSS realm. and at $100 per developer seat, the cost of entry is really not that high.
This comment is fully compliant with RFC 527.
Croquet. There's nothing else like it.
RealmForge is a really good-sounding option for what you're talking about. It's an LGPL project. It's basically a game development kit built on top of Axiom (the C# version of OGRE). It's cross-platform, and one of its main advantages is the ease of content creation. It's still in development, but I must say that the framework is incredibly well-thought out, and I can't wait to see where this thing goes.
FreeWRL
You won't be able to do the type of things that an engine like crystal space would enable, but you will have easy access to web-like hyperlinking and information presentation. It depends on how game-like you envision the final product being and what information you are trying to convey.
now open source./
But it was an RPG for little kids to learn math. Basically they start in a town, that teaches them the basics about math and pointers about that section. Then they walk around the wilderness, being challenged with math problems. Eventually they face a boss that gives them several math problems they have to solve. When they complete an area, they move on to the next. Maybe have them choose their adventuring buddies that may give them hints in math-battles. It'd works because the way math builds on past math foundations. Start with counting, and if you put enough levels in, you could have them up to calculus.
God spoke to me.
2 AMAZING 3D game engines developed in Java. this makes any environment that is built cross platform on Win32, OSX and Linux straight away!
http://www.agency9.com/
http://www.jmonkeyengine.com/
Garage Games has an open-source 3D engine that you can license for $100. You can even sell the final product (up to $250,000 worth of it) without paying a dime more.
"I have never won a debate with an ignorant person." -Ali ibn Abi Talib
What about OGRE? Not used it but it looks like a competitor to CrystalSpace.
You are not a brain: http://books.google.com/books?id=2oV61CeDx-YC
It depends what you want.
If you want to make it easy I'd recommend Torque from GarageGames. It has greate documentation, lots of premade widgets, it's also finished and being used for a number of games. Mainly it has the polish that makes working with it enjoyable. (Tutorials, prebuild samples, etc) Not open source but has a similar community
Someone else has already recommened Crystal Space. If your looking for an example of an MMORG check out PlaneShift.
Another engine that I've been impressed with is ORGE
Anyway as I stated at the begining if you are trying to get something done ASAP go with Torque it has a lot more support. On the other hand if you want something free try Crystal Space or ORGE. One last project you might find interesting is WorldForge they have been building MMORG for a while now. They still haven't finished anything but it's more like a backend and might be useful.
"My research advisor wants to create an online, collaborative science education game for middle- to high-school students; we need a 3D engine! A pioneer in internet-assisted laboratories (iLabs), he has developed real experiments that students can operate remotely over the internet."
http://www.curl.com/solutions/demos.php
Check out the graphics demos.
How about TrueVision3D? I have tested many free/low cost engines and this one is the fastest I have used. Its free for noncommercial use and only costs $150 for a normal license. Check it out at www.truevision3d.com . It has very fast animated model rendering, fast meshes, fast terrain...everything you need to make your MMO. The 6.5 version is in beta and is lightyears ahead of the currently available 6.2 so in a few months you will have access to a fully optimized DX9 engine.
They are also writing a net engine to handle multiplayer which is being designed with MMOs in mind using technology that they have developed in their professional careers so its tried and true networking. Check it out, its free so you have nothing to lose.
Quake2 is GPL'ed
... well you get the idea. With a bit of searching
Quake3 will probably be in the near future, and has a very active community of modders/tools/information/...
Cube (and nextgen Sauerbraten) are zlib licensed:
http://wouter.fov120.com/cube/
fun for its realtime ingame editor iirc
Ogre is LGPL'ed and also very active community
http://www.ogre3d.org/
I think you must be able to find some decent open source stuff.
I'm not sure creating an MMOG to teach students will actually teach them science. What will most likely happen is 1 student will figure everything out and then post it online and everyone else will copy-paste his directions and not learn anything, just like it is with most quests in MMORPGs. Also how would you pull them away from games like: WoW, GWs, CS:S, Halo 2 etc. and get them to play an educational game? The only way I could see getting them to play it would be to have teachers integrate it into their curriculum but there's still the 1st probelm
There are 11 types of people, those who know unary and those who don't.
This sounds very, very similar to a project idea that I've had boiling in the back of my mind, and I would very much love to be a part of this. Can you let me know what school you're at, and who your research advisor is?
The headline says "MMOG", but the blurb reads as if they may only be a few players together within an instance. Similarly, it's unclear whether the game-world is to be a continuous, load-as-needed sprawling environment or if discrete levels would suffice.
These are significant aspects of a game engine's architecture and you are not going to want to have to customize them in; clarifying the requirements here could significantly reduce the number of possible solutions. (Although, given the average slashdotter's reading comprehension and ability to stay-on-topic, it probably wouldn't make much difference in the responses)
OGRE and already mentioned Crystal Space are two most popular open sourced 3d engines, but OGRE community seems more active lately. Here is a summary of OGRE vs Crystal Space
The University of Nottingham has been working for years on MASSIVE which is designed explicitly for this purpose. Prof. Benton and Dr. Greenhal have been working on this for years. Last time I experienced it the graphics were more 1980s VR than Doom 3, but you were able to manipulate the environment collaborativly (I built a house with 10 other people) in real time, use it as a meeting space with full audio and very low lag. Not sure about the licensing, you would have to ask them.
Scared of flying, pointy things snce 1979!
First off, no matter how much money you have, stay far far away from the Torque3D engine. It's antiquated, bulky, and not designed for an MMOG setting. I know this from first hand experience as part of the development team of Virtual Canada. Tourque3D was the engine behind it, and we could only use 64x64 textures, no stretching, very limited size in maps, and well, since we modeled 15 major canadian cities and natural wonders (ie: Niagara Falls), the Tourque 3D engine just collapsed under its own weight. So, we end up with a game that looks and feels 10 years old, yet requires a modern day computer. It was barely able to handle processing a complex map layout like Ottawa, which since this was a federal government project, was the most detailed map, when there were just 50 connections running around. The program looks worse than the first Half Life, but required a minimum 1ghz cpu, go figure.
How you would like for students to interact within the system. What elements do they gather data to make their inferences, and what elements aren't important? Does the system have to change in response to their tinkering? Do they work in teams or pairs? If so, what do you need within the system to allow this to happen? And -- HOW WILL STUDENTS BE ASSESSED?
// http://www.projectkir.org/
Certain educational MUVEs capture data about student involvement within the environment; for example, where they went, who they spoke to, etc. Think about what it is you want to measure, and how you will measure that. Consider how children will learn -- be very precise, and don't expect too much! Remember that class sessions are short, and technology in schools unreliable. Design for -- as Chris Dede likes to put it -- "defenestration".
Finally, when you think you have a system in place, evaluate it early and often! It's tough to build an educational environment, and tougher still to build a good educational environment. There are many who still don't believe MUVEs can teach the explicit skills we might want students to learn... and they might be... could be... correct! So, before embarking on such a huge project, (please) read the literature, and best of luck!
--dave crusoe
Another option for the engine could be Panda 3D.
http://panda3d.org/
(used for disney's toontown online)
I was talking to Jesse Schell, the chairperson of the IGDA, about two weeks ago at the RPI GameFest about the work he's been doing at CMU. We were using Torque for our game development but he actually told me he was using Panda3D for their work. Not surprising when you consider that he's a former employee of Disney but apparently it has great qualities including editing, compiling, and modifying the engine while it's running (I believe it has some Python hooks or somesuch). Anyways, it hasn't been mentioned here so I thought I'd drop its name.
Thanks for your answers, all. I was able to present my advisor with a pretty comprehensive list of alternatives. And I think I am hooked on Ask Slashdot for life...