Hiya,
Re your new patch...
-#define IWineD3DDevice_CreateVertexBuffer(p,a,b,c,d,e,f) (p)->lpVtbl->CreateVertexBuffer(p,a,b,c,d,e,f) +#define IWineD3DDevice_SetFVF(p,a) (p)->lpVtbl->SetFVF(p,a) +#define IWineD3DDevice_GetFVF(p,a) (p)->lpVtbl->GetFVF(p,a) +#define IWineD3DDevice_CreateVertexDeclaration(p,a,r) (p)->lpVtbl->CreateVertexDeclaration(p,a,r) +#define IWineD3DDevice_SetVertexDeclaration(p,a) (p)->lpVtbl->SetVertexDeclaration(p,a) +#define IWineD3DDevice_GetVertexDeclaration(p,a) (p)->lpVtbl->GetVertexDeclaration(p,a) +#define IWineD3DDevice_CreateVertexShader(p,a,r) (p)->lpVtbl->CreateVertexShader(p,a,r) +#define IWineD3DDevice_SetVertexShader(p,a) (p)->lpVtbl->SetVertexShader(p,a) +#define IWineD3DDevice_GetVertexShader(p,a) (p)->lpVtbl->GetVertexShader(p,a) +#define IWineD3DDevice_CreatePixelShader(p,a,r) (p)->lpVtbl->CreatePixelShader(p,a,r) +#define IWineD3DDevice_SetPixelShader(p,a) (p)->lpVtbl->SetPixelShader(p,a) +#define IWineD3DDevice_GetPixelShader(p,a) (p)->lpVtbl->GetPixelShader(p,a) +#define IWineD3DDevice_CreateVertexBuffer(p,a,b,c,d,e,f) (p)->lpVtbl->CreateVertexBuffer(p,a,b,c,d,e,f) +#define IWineD3DDevice_BindVertexBuffer(p,a) (p)->lpVtbl->BindVertexBuffer(p,a) +#define IWineD3DDevice_UnbindVertexBuffer(p,a) (p)->lpVtbl->UnbindVertexBuffer(p,a)
Some of these are defining functions which dont exist yet in wined3d, which will get confusing. Additionally, the functionality for the wined3d vertex buffer is almost untested, because its just copying code and ensuring it gets driven. The d3d9 -> wined3d support isnt enough to prove your changes actually work, and I would rather we didnt put in untested code during the migration phase.
I'm finding it hard to keep track of where I am on the migration from d3d8 -> wined3d, so could I ask that you hold off this patch for a month or so, until I can get enought support into the wined3d library so you can prove the new code works. I could do with the d3d8, d3d9 and wined3d code staying fixed as I try to shuffle it all around.
Once I get enough in wined3d to be able to prove parts of it work, and we are using it from both d3d8 and d3d9 then by all means put in this change. I'd also comment that this change is desperately needed as hardware buffer support should speed up the d3d support too.
It seems odd to turn away new code, especially when hardly anyone else helps on the d3d side! Jason
On Thursday 21 October 2004 16:34, Ann and Jason Edmeades wrote:
Hiya,
Hi Jason :)
Re your new patch...
<snip>
Some of these are defining functions which dont exist yet in wined3d, which will get confusing.
Well, it's only preparative work. I'll try to help you doing this "boring" part :)
Additionally, the functionality for the wined3d vertex buffer is almost untested, because its just copying code and ensuring it gets driven.
I have tested it in a separate code. But i thought that including it directly into wined3d instead of d3d8 would be better.
The d3d9 -> wined3d support isnt enough to prove your changes actually work, and I would rather we didnt put in untested code during the migration phase. I'm finding it hard to keep track of where I am on the migration from d3d8 -> wined3d, so could I ask that you hold off this patch for a month or so, until I can get enought support into the wined3d library so you can prove the new code works. I could do with the d3d8, d3d9 and wined3d code staying fixed as I try to shuffle it all around.
As you want, i'll continue playing with my modified d3d8/d3d9 waiting you to complete wined3d migration :) Alexandre, i can forget my patch
Once I get enough in wined3d to be able to prove parts of it work, and we are using it from both d3d8 and d3d9 then by all means put in this change. I'd also comment that this change is desperately needed as hardware buffer support should speed up the d3d support too.
It seems odd to turn away new code, especially when hardly anyone else helps on the d3d side!
Jason
Thx
Best Regards, Raphael
Well, it's only preparative work. I'll try to help you doing this "boring" part :)
Yeah - its painful going, but I'm doing it slowly both so I can be careful to cross reference everything with the MSDN, but I'm also trying to keep patches to a reasonable size and get them committed before I get to the next chunk.
I have tested it in a separate code. But i thought that including it
directly
into wined3d instead of d3d8 would be better.
Sorry - I didnt mean to imply it wasnt tested as such, as I guessed (and appreciated) the way you had done it. The 'problem' I have is when I change the d3d8 code to use wined3d I would rather it was in the same state as it is now, so regressions become obvious.
As you want, i'll continue playing with my modified d3d8/d3d9 waiting you to complete wined3d migration :) Alexandre, i can forget my patch
I think as soon as the device interface of d3d8 is modified to fully use wined3d, then anyone can start changing things. I'll give you a shout anyway, but it should be done in a month or 2 hopefully (but dont hold me to it!)
Jason
-
Ann and Jason Edmeades -
Raphael