Stable enough for external developers?

[polyfractal]

I've been shopping around for a MMO framework to use as I figured preexisting technology would be quicker/easier/more stable than something I could create. Since this is just a hobby project of mine, I don't enjoy the prospect of creating an entire MMO framework Multiverse3D looks very interesting and powerful. I haven't had a chance to download the sources yet so take my questions with a grain of salt.

In your opinion as the lead developer, how stable do you feel MV3D is for use by an external developer? I realize it is far from feature complete and will likely undergo a lot of changes. Is there enough core functionality to support development of a game at this point in time? Of course, if I were to begin developing my own project on MV3D I would also help fill tickets on the main MV3D branch as it is a mutually beneficial arrangement.

[SirGolan]

That's definitely a tough question, and the answer depends a lot on what you want out of the framework. With the exception of some examples, I generally have been keeping game related things out of MV3D itself. This is part of the planned direction of the framework, so it likely won't change. I envision someone (if not me) will write modules separate from the core MV3D code that include game logic, standard NPC stuff, etc; however, none of that is currently done or available. On the other hand though, I've been trying to get Glyph excited about the prospect of MV3D & Imaginary, which I think would be a great way to incorporate game logic into MV3D. I wouldn't be surprised to see that happen in the near future, but again, it's not in place now.

Back to the core of your question. Yes, I do believe that MV3D is stable enough for external developers. There are only a couple of changes I'm thinking of making that could effect external developers. One of those is moving some things over to being stored directly via Axiom. I'd hopefully be able to provide some sort of migration path there, but also, most of the stuff I'm moving over would only be an issue for you if you are planning to make your game isolated from other MV3D worlds (i.e. if you aren't part of the single sign on described here).

I have definitely been looking to get more people to start using MV3D, so it would be great to have you on board. Bug report tickets would be excellent. Just let me know your Trac login and I'll give you permissions to make tickets and such. I think that the more people who are actively developing games or simulations using MV3D, the clearer picture I can get of what I need to concentrate on next.

Thanks,

Mike

[polyfractal]

Well that is certainly encouraging! I'll make a Trac account and get a copy of the source to start playing with. I've still got some more work to do on my design document before I start coding in earnest however.

If you don't mind, I have a few questions regarding MV3D. Since it looks like MV3D is designed primarily to be a framework/platform, I'm guess it is relatively agnostic to the gameplay of individual servers? Ie. would it be possible to have a radically different setup (RTS, for instance) instead of the traditional RPG? Similarly, how capable is the network code at handling high-volume like one encountered in an FPS?

(If your curious, I'm debating between two game designs that I've been working on for a while, an MMOFPS and an MMORTS )

[SirGolan]

You've got it correct. MV3D aims to be a fairly general purpose virtual world simulation framework. Your world doesn't even have to be 3D to be honest. I keep meaning to build some sort of space demo world and a 2D world to show off those features. As far as an MMORTS, I think it should be no problem and probably wouldn't require much tweaking of existing code at all. You'd definitely have to write some things to handle the difference in control (i.e. maniuplating multiple units instead of just controlling one), but I can think of a couple of fairly straight forward ways to do that right off the top of my head.

An MMOFPS could be a little more difficult in some ways. Depends on how comfortable you are with UDP. Right now, MV3D uses TCP for everything. It honestly wouldn't be too hard to abstract out the position/rotation updates to UDP, I just haven't had the time (plus TCP has been good enough so far). The updates are already fairly self contained (mv3d.phys.StringFlinger / mv3d.phys.StringCatcher), so really it should be a matter of a new implementation of those classes that used UDP instead of strings. The other thing that is completely lacking currently is lag compensation. This one is starting to get on my nerves, so I may do it soon, though I don't have it on the timeline yet.

Hope that answers your questions.

Mike