PureBasic Forums - English

The minimum target platform is a geforce 2 ... so I can do a single pass with 2 textures at a minimum in hardware. Lightmaps are supported, curves ? not sure what you mean by this. Shadows are done in the stencil buffer, so they will not be a problem on your system.

If your system cannot do single passes then I will have to do multipasses, which means the object will be rewritten twice, which takes a lot of cpu, gpu power, if you have lots of objects using multiple textures. Kind of confusing, but I am trying to make it as non-transparent as possible for everybody so you don't have to worry about those things.

Also a sprite engine will be included for doing user interfaces or even 2d games...

Damn, ok nvm then, Im no longer interested in the project lol. Good luck with it. I believe it will be a great product for PB users.

Damn, ok nvm then, Im no longer interested in the project lol. Good luck with it. I believe it will be a great product for PB users.

Why the loss of interest ?

Hmm, Here is my final tought about your engine.

When I consider all consequences I think that here in internet have many good engines to use with PB. In fact I cannot see why your engine sould be better then 3DSate's in example.

http://www.3dstate.com

Or OGRE. You will spend a lot of hours in project which is judged to nonexistance because of timing. In my opinion your engine will be ready about year or two, and I think many engines will be improved at point of 3d realistic worlds, when your engine will just render BSP scenes, and in better work collisioning system.


I don't know man, if you gona make it, who's gona use it?
I'm surely think i'm not.

When I consider all consequences I think that here in internet have many good engines to use with PB. In fact I cannot see why your engine sould be better then 3DSate's in example.

Well a real 3d engine cost over $50,000. I should know, here's a contract for the evaluation license of the Morrowind Engine my old company secured. Not an easy task at all, I should not be posting this at all, but seeing as my old company is dissolved, I don't mind. Even though it is unprofessional.

Hey codemonger,

What is a rough estimate when your engine will be able to be bought? And how do you think it will compare to pb's 3d engine by then? Since rumor has it that release 3.90 will add alot to the 3d engine. And if will you, if the release adds affects the 3d engine, will you be using any of it or are you going to be using pure directx?

By the way I think you should finish no matter what. It will be a good learning experience, after all who ever day gets to write a 3d engine straight from directx or has the knowledge and time to do so?

Thanks Master Games,

I don't have a date for purchasing or anything to that degree yet. But I will soon set up a BETA testing stage.

I don't see PB as a 3D engine platform natively. Personally I think it is not good for PB to add 3D support natively as it only limits it's use as a pure programming language, especially when as users we want requests and changes, well OGRE is seperate so chances are we can't get those changes or requests.

From a professional standpoint, I don't think adding in an opensource 3D engine is a good idea, who knows where OGRE will be in a year or even two or three, or how hacked the code is or will be.

IBASIC is a good example, where now their new publisher has seperated the 3D from the core engine as a seperate product. Why ? Because it is a seperate product.

Anyways the CMG3D engine supports 100% DirectX and the code is very tight and organiszed. I am a speed freak and an organisation freak, so these are bonuses in my coding. I hate cutting corners. Oh, I wish I knew Assembly or had the time to learn it

Anyway I will keep everyone posted. Right now this second, I am finishing up a dll importer for importing commands and using them natively from the CMG3D IDE. So you can easily import dll functions and not have to use callfunction or callfunctionfast.

My daughter is sleeping over at her friends so, it is peace and quite tonight. Programming Time! Anyway if you have any questions feel free to ask. And remember everything is in the development stages, so their is no need for anyone to rush to judgement, yet. At least not until some screenshots get posted.

Thanks

Thank you for the quick reply. I just want to say my biggest complaint with the ogre engine is that they made their own model format. I like engines that use the standard format, unless they make it very easy to convert (like making a program to convert the normal model formats (.3d, .x, ...) making plugins, makes the user to have to have the program that the plugin was made for, and not ever one has them.

After all we are not fithy rich (atleast I know I am not).

By the way when do you think you will be able to release some demos made with your engine for us to see its power and speed?

Too early right now for demo tests for speed and visual effects. I here you about the format ... the most generic and free format is the x file format. And you know it's format is going to be compatible, extracting information from loaded .X is easy. It will probably be the format of choice. But that is something I haven't yet focused on. If you have any ideas feel free to throw them my way.

> The minimum target platform is a geforce 2

the most spread low cost card is the geforce2 mx, though that is *not* a geforce2

re format: might add a little code to do model conversion, cmg3d <> ogre

i'm currently helping somebody else developping an engine in c++, doing the testing for a future game

one of the reasons he is developping his own engine is the need for outdoor scenes: lots and loads of polygon reduction...

re curves: shadows, curves and lightmaps are mostly needed to make things sexy

Codemonger wrote:

Damn, ok nvm then, Im no longer interested in the project lol. Good luck with it. I believe it will be a great product for PB users.

Why the loss of interest ?

I have an original Geforce 2 mx. Not the crappy junk dubbed mx 200 or mx 400. I have the real deal. I find that alot of engines that aim for GF2, does not work or work very slowly on a real gf mx.

So if your aiming your engine for a geforce 2, there' really no way I can use it.

Oh no, your a perfect candidate for testing.. Everything will run fine on a TNT32 Card, but the target is a geforce2 which i have in my PC right now. I picked up a couple of video cards a while back for testing for a commercial game. No use for them now, but they are great for testing. As far as speed etc...

As far as effects are concerned like dot3 bumpmapping, your card will support it, and should be able to get it done in a single pass with 2 textures. Don't want to get into detail right now. But your OK, I am not cutting out anthying below geforce 2, it just what I have available as far as hardware right now, and makes a good benchmark platform. Hope you'll help out for Beta testing eventually, as your card would be good reference.

Scared the begezus out of me Then I am definately interested.. big time I currently have the Torque Engine (torque.com) and am fiddling around with it... not much to compare atm Anyways, count me in, when you need testers I can give my complete system specs, if it'll help

Thanks Shanara, everybit will help. I bought torque some time ago and was very unhappy with it. It is very fast, but a little old. The problem I have is that the source code is poorly written, looks more like hacked code. Something like 400000 lines of code ... Then I bought Cipher engine and all the code is properly documented and very professional, but I think they made the mistake of using a specific layout for creating a game, sort of it's own pipeline which is not very flexible. I could never get used to it, it is very annoying to have limitations.

In between I bought Blitz Basic, and it is overall easiest to use, well documented, well structured. But when it came time to number crunch and do any newer stuff, it fails drastically unless you have a super machine. I was at work and downloaded a cool demo of a game someone was making and tried to run it on an 800 MHZ computer ... well I got about 1 FPS ... not impressed... The game did not have any special effects that would cause this, just Blitz could not handle high capacity environments, or just a bad programmer.

So I am targeting primarily speed ... This would allow for outdoor terrain systems, many many world objects, like thousands of trees etc ... I have in my bookmark section links to every topic imaginable, and have read countless articles and papers. I also read and re-read Nvidia Tech articles on optimizing speed, I take notes vigorously as if I am in school :roll: It's a bad programmer habit I picked up years ago. When I wan't to learn something I read as much as possible and take notes the second time around reading. This is how I learnt DirectX7 years ago, I forced myself to read the Microsoft DX Manual over and over again. It really does makes sense, but helps if someone explains it in english.

Not to forget I have experience at the helm programming 3D and optimizing to use the video card to it's full extent. But DirectX9 is more forgiving and does more managment than 8 or 7, which is very nice. So I can focus on other things. Anyways I'm just blabbing right now, time to get back to coding.