19 replies to this topic

[idreamlovey]

hey,
i am abt to complete my first game based on car racing. My engine is based on directx 9. It took me long time to complete it and i wil lauch it in march.
Tel me whether it wil stands with other games this year or not cause i worked so hard for it and feeling somehow insecured as i basically used fixed pipeline which i heard that is being replaced by directx 10.
Also i don't have window vista. Tell me my game will run on vista system or not.

[idreamlovey]

I know its COM object and it wil definetly run on vista system too. But tel me whether u people started working on directx 10. I mean Do Current Engine get started of switched to DirectX 10 ?
U know first time, i m feeling thrill of it.

[Reedbeta]

Vista has just been released, so people are only going to be starting DirectX 10 development now (or they may have started while Vista was in beta). In any case, DirectX 9 and the engines built on it will still be around for a long time.

[Spodi]

If you can make your product fine in any previous version of DirectX, upgrading isn't needed. DirectX has always had backwards compatibility. Theres plenty of DX7 games out there that people still play, just since they don't "need" to upgrade.

[dave_]

You're asking the wrong questions. DX10 will just be an extension of DX9, so of course it will surpass it. What would be the point if it didnt.

[.oisyn]

Actually the D3D10 API changed very drastically with respect to DX9. Of course DX9 is still supported in Vista, but that has got nothing to do with D3D10

[TheNut]

DX10 has at least 2 more years before it will really kick off for mainstream use, and even then it doesn't make it better. Microsoft made a lot of assumptions with DX10 and only time will tell whether or not those assumptions were valid. DX9 with HLSL or OGL with GLSL is still sufficient for today's market. The quality you can pull out with either of those two setups is still enough to bring awe and amazement to a gamer's heart.

[dave_]

Although the API will change drastically functionally its not that different. To start with you'll probably be able to do everything you can with D3D10 with D3D9

[.oisyn]

Euh, don't you mean that the other way around?

[dave_]

No, I don't. Although Shader Model 4 is only available in DX10. Its just a different level of abstraction. It might be easier to write things for DX10, but that doesnt mean its not possible in DX9.

Show me an example of a DX10 demo that would not have been possible under DX9.

[.oisyn]

What about the geometry shader, memory exports, texture views, etc.

[TheNut]

As someone from Microsoft said, it's all about "Doing more". It's about the features and performance that differentiate DX10 from its predecessors. Extra hardware accelerated features not presently available for DX9, such as geometry shaders is a big plus. Visually speaking, DX10 has a better "price per performance" ratio than DX9, so DX10 demos can look visually better than DX9 under the same frame rate constraints.

In either case, MS said "Let there be DX10", and ultimately the market will have no choice but to upgrade to it, for better or worse. Until it hits mainstream, enjoy DX9 or OGL for the time being. SM3 is still quite powerful.

[GhostManZero]

They'll be making a new OpenGL Version for the SM4-capable cards soon, the first new version in over a decade and a half, as they say, it wont be just a bunch of extensions.
So, if we're lucky, D3D10 will have some competition! :D
I myself cant wait to see all the features they'll implement on D3D10 and the new version of OpenGL. Should be quite interesting! :)

[.oisyn]

GhostManZero said:

They'll be making a new OpenGL Version for the SM4-capable cards soon, the first new version in over a decade and a half, as they say, it wont be just a bunch of extensions.

Who is "they" and what happened to OpenGL 2.0?

[GhostManZero]

They is the OpenGL ARB Group, and THIS explains it all: http://www.opengl.or...ticle/vol002_1/

OpenGL 2.0 and "before" was implemented through extensions, they're going to build a completely new API for SM4 so it can be considered a true "new version" since they're actually changing the API and not just setting it as some new extensions, that's what i meant at least.

Quote

The main task for the ARB is to deliver two new OpenGL releases in 2007. The first one, code named OpenGL "Longs Peak" (the actual releases will have version numbers), is slated to be released in summer 2007. The second one, code named OpenGL "Mt. Evans", is targeted for an October 2007 release. Why code names? We want to give the ARB's marketing group a chance to think through what the right names would be for these releases. Too many suggestions have already been made, including OpenGL 2.2, OpenGL 3.0, OpenGL 3.1 and even OpenGL 4.0. This is not the time yet to pin down the version number, and therefore we'll be using code names.

OpenGL Longs Peak will be a significant departure for us. While there will still be backwards API compatibility, the new "Lean and Mean" profile, and a substantial refactoring in terms of the new object model, make it in many ways an entirely new API design. This is an ambitious task and requires a high degree of commitment from the ARB members. We are already seeing some welcome participation from Khronos members who were not members of the old ARB, and hope to see much more.

While OpenGL Longs Peak will be implementable on current and last generation hardware, OpenGL Mt. Evans will only be implementable on the newest hardware available. The OpenGL Mt. Evans release will be a continuation of OpenGL Longs Peak, with a lot of new functionality added. Some of the highlights are: geometry shading, a more central role for buffer objects, and a full integer pipeline accessible via the OpenGL Shading Language.

[.oisyn]

I know what you meant, but I thought OpenGL 2.0 did that already . I did not know it was, again, based on extensions.

[GhostManZero]

Yeah, but OpenGL 2.0 is still extension based, but since they'll actually redesign the API i consider it a real "new version", that's all :D

We all know extensions suck 'cos they usually behave differently from video card developer to video card developer, that's why GLSL sucks on most ATI cards, but rocks on nVidia cards (at least in my opinion, i keep getting incompatibilities when using ARB GLSL extensions to compile GLSL code that was working on an nVidia card, but was able to fix it.

So a redesign will probably do wonders! :D

[.oisyn]

Well, yes, but it does have advantages too.

For Tomb Raider: Legend, we were having problems with nVidia's shader optimization code under D3D9, causing massive stalls whenever new shaders were loaded (and we used a LOT of them). To circumvent this, they released a new driver and we had to create a special texture in order to disable that feature.

Another one is nVidia's hardware shadowmap feature. In OpenGL, it's nicely exposed through an extension. On D3D, you just have to sample from a depth texture (which usually isn't allowed), which triggers the hardware shadowmap feature in the driver.

Sometimes, extensions can be nice. I do like D3D10's approach, though (and I guess the new OpenGL version). Just: everything is there, all features are supported, no caps checking, do your thing, it'll work.

[GhostManZero]

Yes, it'll be quite a good help! :D
I seriously hope D3D10 will have some good competition from OpenGL! ^^

[rouncer]

displacement mapping with directx9.0 is alot harder to implement without
the geometry shader.
its like trying to make a splatted terrain with the fixed function, or bump
mapping with the fixed function.
making a splatted terrain without a shader is a complete pain involving
texture stages compared to simply adding weight factors together.
displacement mapping in dx9?
youd have to store multiple mip levels of vertices for all the geometry in the
level, so youd run out of ram before you could do anything.
the geometry shader tesselates it on the fly.

but you can do a displaced terrain with dx9, as long as you are on shader 3.0
that supports floating point textures, which is what the vertex shader uses
to perturb the vertices.

also another thing in directx10, is having access to the actual id's of the vertices
and triangles, which is so useful and was a real neglect not having it.

i cant believe how many times i wanted them.

directx 9, in my opinion is going to be a lot more of a running goose chase
(or was) to program with compared to dx10.

and ive just got my new dx10 video card and new dual core athlon. awesome.