Let's Make Our Own Fps!


-Tj-

Active Member
Joined
Sep 2, 2009
Messages
842
There has been a lot of talk of this stuff: Source on Pandora? Rage on Pandora? UE3 on Pandora? Etc. Of course that's not all accurate, but my point is there's a lot of this "can we put this [closed source engine] on Pandora?" stuff going on.

Which begs the question, can't we just, you know... make our own engine for Pandora? :) Surely there are enough competent coders who have ordered or already have their Pandora who'd be able to code something up for our system? I know it'd be a lot of work to do something like this so I know it's kind of a stretch, but there's also a bunch of open source 3D game engines out there... Perhaps we could start with one of those, optimize it for Pandora's average (non-overclocked) specs, then start building/modding?

The way I'm thinking is this. If we can put together a decent enough engine, we could probably get to a point where we could recreate most of the FPS that have been talked about, to a certain degree. I'm not talking spot-on (that'd be impossible), but a reasonable facsimile, much like how modders have recreated levels, weapons, and/or characters for other game engines.

Most of my freelance work is nearing completion now and I would be interested in making assets for something like this after I'm done with it all. It would also help me keep creative outside of work.

Whaddaya guys think? There's a lot more I could write on this subject, but for now I'm just interested in how many people (users and/or developers) would be interested in something like this.

I've also posted this on the official boards in case I've missed some people. :)
http://boards.openpandora.org/index.php?/topic/158-lets-make-our-own-fps/
 
Well, work with the OpenGL ES on board the Pandora is still fairly pioneering at this stage. I'm guessing an engine couldn't be built until that is understood better.

The reason why there's so much talk of porting other engines is simple: "Why re-invent the wheel?"

There is a stupid amount of work involved in creating a good 3D graphics engine from scratch, let alone creating the other elements required for a proper game such as models, levels, sound and AI. I'd be very surprised to find many people who were willing to put in so much effort without financial compensation.

Sorry to pour cold water on your idea, but porting an existing engine would typically be about 100 to 1000 times quicker than building one from scratch. That being said, if you think you've found an open-source 3D engine that could be ported to Pandora then I'm sure there'll be some people that are all ears.
 
TimmyB said:
There is a stupid amount of work involved in creating a good 3D graphics engine from scratch, let alone creating the other elements required for a proper game such as models, levels, sound and AI. I'd be very surprised to find many people who were willing to put in so much effort without financial compensation.
Well yeah, that's why I said:

I know it'd be a lot of work to do something like this so I know it's kind of a stretch, but there's also a bunch of open source 3D game engines out there... Perhaps we could start with one of those

The way I see it, even if we were to get the source for UE3 or HL2 there would be just as much work to do on the graphics front. Graphics and models would need to be optimized to run well on the Pandora's small screen, which means reducing and/or recombining textures to compensate for the lack of shader power, engine optimization, polygon reduction, which also may mean redoing UV maps and reexporting animations, perhaps even redoing skeletal structures to reduce the number of joints, and so on. Levels and maps would need to be optimized, redoing and rechecking areas you may not even know about. In a way, it might take less time to make new low-res assets than it would to convert/optimize old high-res assets.

What I'm trying to say is there's more to porting a game than just moving it from one platform to another, especially if you want it to work well. What worked for Quake 3 likely won't work for the games with much higher minimum system requirements.

Anyway, there's a video of Irrlricht Engine running on Pandora done by cpasjuste over on YouTube. Have a look if you haven't seen it already: http://www.youtube.com/watch?v=iF42uQwtQ_U. It's not a very clear video, though, but you can see it works. Yes, there's more to a game than just opening examples up and viewing them (AI, collision, effects, music, sounds, etc.) but it's a start.

There's also PixelLight, which supports OpenGL ES 2.0.
 
Last edited by a moderator:
Once Darkplaces (for example Nexuiz/Xonotic uses this engine) is ported, you will be able to create amazing graphics on a fantastic 3D game engine.
 
Nexuiz looks amazing. If that's using the Darkplaces engine, that'd be a fantastic starting point for something we could do. Haven't had time to check Xonotic yet, maybe later tonight I'll have a chance to do that.
 
-Tj- said:
Nexuiz looks amazing. If that's using the Darkplaces engine, that'd be a fantastic starting point for something we could do. Haven't had time to check Xonotic yet, maybe later tonight I'll have a chance to do that.
Xonotic has no release, yet. It is a fork of Nexuiz that was created after a license for Nexuiz was sold to Illfonic who will use it for a commercial game on the Xbox 360 and the PS3. That decision split the community. Xonotic is supposed to be the "new" free and community-driven Nexuiz with new content.
 
Last edited by a moderator:
I'm pretty sure Ogre has (at least the beginnings of) an OpenGLES rendering backend. Working on/improving that would be a good place to start if you want a good game engine for Pandora.
 
I was looking at Ogre as one of the potentials, but as of today it doesn't seem like they have anything OpenGL ES-compatible released yet. Maybe I'm wrong and my Google-fu sucks. :p
 
Apparently they were part of GSoC (Google Summer of Code) 2008, and one idea they put forth for students was (paraphrased) finishing the GLES2.0 rendering system that someone had started and was part of the devel trunk. I'll check the code repo and see if I can find it.

EDIT: Here's their GLES2.0 render system code. I couldn't find anything describing the completeness of the implementation - I'd suggest that someone write on their forums if they're interested in getting Ogre working on Pandora (meaning a dev wants to work on their render system if necessary, not a newbie asking for help).
 
oh plsplspls look into getting the Blender Game Engine to run on the Pandora! It's so convenient, easy and powerful to develop games with the BGE. It might not be the fastest engine but it's accessible to a crowd of people 100 times bigger than any regular stand-alone engine: through logic bricks as well as python scripts everybody can convert an ideo into a game!

I already started a threat about this topic (bringing the Blender player (NOT the full version of Blender) to Pandroa) here: http://www.gp32x.de/board/index.php?/topic/55604-compile-blender-for-pandora%3B-need-help-pls/page__p__898358__fromsearch__1&#entry898358
 
Last edited by a moderator:
I agree with zapman on this one... this is very K.I.S.S. and would, overall, be the quickest to port... I bet it's just as simple as a recompilation... Such is true going from x86 to x64... but a short read of blenders bare necessities would probably be the final judge of this possibility... though I doubt it would not run on arm... come to think about it... I wonder if it is in the Debian Extend packages listed in another thread floating around here that I can't remember the link to... :p

that brings another slightly off-topic concept to mind... deb to pnd conversion scripts... but I digress.
 
There's a BIG difference between an x86->x64 recompilation and an x86->ARM recompilation. For example, if there's any x86 assembly (unlikely), it'll work fine when compiled for x64, but will fail for ARM. There's practically no way to say how easy a port will be until it's finished (or until someone's at least looked through the source code).
 
@ Tempel
Has anyone tried that out yet? I'm not a coder and I don't have my Pandora yet so I don't really know what's going on there... I hope it's good, though. :)

I wouldn't mind seeing Blender's game engine going on Pandora, but tbh I don't have any Blender experience. I'm an XSI/Maya guy here, so I'm hoping for COLLADA support, which is pretty universal across every 3D app that supports it. As far as it being simple to port, I dunno. Again, I'm not a coder, so everything sounds difficult to me, and +1 to Vorpeal's post. I would also guess that it'd be easier to port if it supported GLES (is that an accepted abbreviation?).
 
In general, compiling Blender for Pandora shouldn't be a problem: as can be read on this thread ( http://blenderartists.org/forum/showthread.php?t=155801 ) Ideasman42 has already compiled it for ARM architecture without (m)any changes to the code.

The thing that seems tricky is the part about openGL <-> openGl ES. That topic is also adressed on that thread, would be very very nice if someone with porting skills could have a look. Theoretic suggestions what would have to be done are also given.
 
sebt3 said:
If Blender GE was just a simple recompile away, I would have pnd this a long ago...

Have you tried? In what problems did you run?
 
Last edited by a moderator:
I know that it's probably (very!) to mention this, but I can write a pretty interesting storyline if anybody's interested, particularly for a sci-fi type game. Something.. Halo-ish.
Please be forewarned, however, that if your game has prodigious levels of blood/gore, I simply shan't participate. It's nothing personal, it's just one of those things I'd rather not do. Understand :) ?
On another note, I need to stop being lazy and learn Blender. It's just that whenever I look at it, I feel like I just got beamed aboard the starship Enterprise :lol: .
I'd be thrilled, though, if Blender could be used on the Pandora. I'd probably take the time to learn it while traveling.
 
Back
Top