How Cg favors NVIDIA products (at the expense of others)

[demalion]

To SteveG and Geeforcer,

The lack of a quoted reply to some of your posts that are direct repliest to me is because I see nothing stated that I have grounds to refute or counter-argue.

I just maintain that I would prefer assurance rather than depending on things to "work out". It is chiefly my observation of the behavior of humanity in groups where self interest is involved that causes me to lack the confidence you do in your beliefs, though I too hope that is the way things work out in the end, and recognize that the pressures you believe will exist might indeed triumph.

 

[Geeforcer]

But on the grand scale, things have been "working out" so far - why do you think this no longer be the case in the future?

 

[pascal]

I just maintain that I would prefer assurance rather than depending on things to "work out". It is chiefly my observation of the behavior of humanity in groups where self interest is involved that causes me to lack the confidence you do in your beliefs, though I too hope that is the way things work out in the end, and recognize that the pressures you believe will exist might indeed triumph.

Very well said. Is better prevent than try to fix later.

 

[Geeforcer]

Very well said. Is better prevent than try to fix later.

Unless you somehow can foresee the future, this is very, very dangerous and regressive philosophy. Using Glide as an example yet again: Most of us would agree that proprietary APIs are bad. But if you were to prevent the adaptation of Glide in 96/97 on the grounds of "better prevent rather then fix", you would have done a disservice to the development of 3D software and the industry in general.

 

[pascal]

It is not a philosophy. In this specific case is better prevent rather then fix

 

[Geeforcer]

It is not a philosophy. In this specific case is better prevent rather then fix

What makes this case so "specific"?

 

[KimB]

It is not a philosophy. In this specific case is better prevent rather then fix

So we shouldn't have the choice of using Cg because nVidia may try to go back on their current claims of trying to make it an open standard?

 

[pascal]

It is not a philosophy. In this specific case is better prevent rather then fix

What makes this case so "specific"?

Many things:
-We have much more time than in 96/97
-The installed base is the key factor
-There are alternatives coming like the ARB´s HLSL

So we shouldn't have the choice of using Cg because nVidia may try to go back on their current claims of trying to make it an open standard?

Nobody said that. I said we should have the choice of a non proprietary open standard
Maybe you as a declared nVidia fan dont want a non proprietary standard

 

[KimB]

As long as nVidia delivers on what it is currently promising with Cg, it will effectively be open.

Regardless, the one thing that I want out of Cg is a language that will work both with OpenGL and DirectX. That's the really exciting possibility that Cg poses.

Hopefully Cg will evolve enough before its final release in the next couple of months to merge with both DX HLSL and GL2 HLSL. That would be an ideal scenario.

 

[pascal]

I still would rather see a non proprietary open standard, and please stop doing marketing for nvidia.

 

[KimB]

Why is Cg proprietary? It is meant to output industry-standard shaders, and any vendor can make their own shader output if they choose.

And, as others have said, it's a language...it's rather hard to make the language itself proprietary.

Regardless, the #1 advantage that Cg has over the competition for use in games is that it can output DX8 shaders.

 

[pascal]

If HLSL:

-is just a simple language
-the real work is inside the hardware and algorithms used
-has impact in the development and maintenance cost of real-time applications
-will have its use limited by the installed base
-dont need to be done tomorrow
-has alternatives coming
-has some possible implications about trademark, marketing and evolution control

then IMH layman/gamer/consumer/oldman opinion is better have a non proprietary open standard.

 

[Geeforcer]

It is not a philosophy. In this specific case is better prevent rather then fix

What makes this case so "specific"?

Many things:
-We have much more time than in 96/97
-The installed base is the key factor
-There are alternatives coming like the ARB´s HLSL

Yes, and? How does any of this indicate that the market will not sort it out? Do you realize that what you advocating is anti-competitive? The age-old axiom still applies: More competition is good. Do you, pascal, have better understanding of what is "good for the industry" then the industry itself? I have much more faith in the ability of the market to sort things out then in armchair analysts like yourself dictating in advance what is good, what it bad, what should be developed and what should not. If humanity listed to people like you, who'd rather have a product killed off before it could enter the maker, because it is potentially dangerous and "better prevent then fix", we would still be riding horses.

 

[Basic]

swizzling of the different components of the quartet. (Making RGBA to BGRA).

Heh, interesting. No, it doesn't take multiple instructions. I guess you can say it is native as well. What you heard was FUD.

No it wasn't FUD, but it was a misstake. He probably heard it from me (in a post that weren't supposed to get here). And it was shortly thereafter retracted when it was clear that DX9 was meant to have that function since the beginning. My misstake were based on ATi not talking about the added swizzles in their R300 info. But I say again, that doesn't mean that it isn't there, there is no reason to doubt that R300 can do full speed swizzling.

 

[Geeforcer]

I still would rather see a non proprietary open standard, and please stop doing marketing for nvidia.

You do know how OGL started out, right? Learn from the history.

 

[Doomtrooper]

CG is a language, not a API..those comparisons are moot. OpenGl on its debut didn't have two industry standards being released 2-6 months later either.

 

[pascal]

Geforcer

I am not dictating anything, I am just expressing my opinion, the industry will decide what will be done.
The anti-competitive practices come from large companies dictating to the entire world what will be done.

You are gratuituslly insulting me saying that I am an "armchair analyst" and saying that "people like you...".

 

[Doomtrooper]

There is a few people here that have had business experience...people that have open minds besides a few slams from the pro CG group including Reverend. Joe, Pascal, Shark, Noko and a few others see it the way I see it...its a move to put Nvidia in the drivers seat...I can see no answer to why else we need three HLSL...if the effort to design a entire new language by one IHV isn't a sign of cornering a market..I don't know what else is. If Nvidia would have put its CG team working towards OGL 2.0 instead of blocking extensions..Opengl would not be in the mess its in now.

 

[Geeforcer]

The anti-competitive practices come from large companies dictating to the entire world what will be done.

You are gratuituslly insulting me saying that I am an "armchair analyst" and saying that "people like you...".

No pascal, anti-competitive propositions can come from everyone who says thing along the lines "we don't need more products" or "2 is enough" - it doesn't matter if you are an individual or a huge corporation.

Let me ask you you this: how many graphics IHVs do you think we need?

I think that "armchair analysis" is a pretty accurate description of what you have said so far. You already deiced how many HSLS we heed as well as what they should be, all that prior to any HSLS being fully available.

 

[Geeforcer]

I can see no answer to why else we need three HLSL...

So DT, do tell us how many HSLSs we need. Why do we need 2, but not 1 or 5?

Its not for people like you, pascal, etc. to decide what we need, what we don't and how much of it we need. (Thank God for that). We have a FREE MARKET for that sort of thing.