I’m developing an AI library specifically for vehicle/character control and pathfinding. I’m hoping to attract game developers who are using Java3D and is in need of this kind of technology. I’m saying Java3D mainly because of the behavior classes that will be included. Other Java game technologies may come later. The programmer may even wish to write their own behavior classes now that I think about it. Anyway, do you guys feel like something you would pay your hard earned money for? Or am I dreaming. I appreciate good honest criticism. I also would appreciate refraining from making insults. Thank you for your time. 
Depends on the price for me. Yes maybe if i was making a game i would pay a littlebit to
get those ready made behavior classes.
If I was just doing a hobby project i wouldnt pay much, if I had a real company then I think I would pay some more.
The AI market is definitely there (just take a look at BioGraphic Technologies) however I’d suggest if you’re going to do this you do it in a supremely generic way. Don’t target Java 3D Game Developers as a market (since they’re really arn’t that many) - make the tools provide generic to any technology and provide example usages within Java3D behaviours (I guess it’s just another level of decoupling/indirection). You’re more likely to hit more of market if the tools arn’t seen or percieved as Java 3D specific.
Kev
I probably wouldn’t pay for something like that… but then again, if I reached a state of despair with my own code I might…
Thanks Zulo! I know what you mean. Price is always a concern and people want to get their money’s worth.
Thanks Kevglass! I have looked into BioGraphic Technologies before and I understand what you are saying about targeting just one technology. I may have to use an incrementalal approach starting with Java3D.
Thanks Aramaz for your reply! I do think there are some game developers that just want to concentrate on the specifics of their games without having to code the physics, AI and things like that. Anything that can cut down their development and testing time would probably be favorable.
You might consider a licence scheme where low revenue productions pay less than big sellers. So maybe licence-costs per sold game/application might be appropriate. There are some libraries (but not in gaming as far as I know) out there, that come with dual-licence GPL/Commercial, so hobbyist can use the library for free.
Regarding GPL: I for myself am not an OpenSource advocat, but I really appreciate to be able to step through library code while debugging…
The idea of hobbyist using it for free sounds interesting. At least as a start to gain popularity.
Anyone else have any comment on this? I would greatly appreciate it.
Okay soem comments:
Behaviors are semi-useless for games UNLESS you design it as a wakeOnFrame(0) behavior or its results can be used truly asynchronously. (If the latter then behaviros are actually good because they will get scheduledi n parallel onsystems with more then one processor.)
I’m thinking you really might be better off designing this as a library with a doAI() entry point and then let the user wrap it ina behavior or whatever else they want including using synchronously in a main loop.
Does this AI do path-finding and motion control? if so then you definitely want to be able to run it synchronous with the other game logic sucha s the game’s collision detection.
That’s the plan Jeff. To do path-finding and/or motion control. Great advice on the behavior idea. Thanks!