OK, another tack since there are still thread participants besides Russ who might not understand my statements.
To which he could reply: that when the same optimizations would be done for games, and provide identical output, and in fact developer relations does exactly this for games, the label "inflating" is completely inaccurate.
To which Futuremark would likely reply: we want our benchmark to show the general ability of your drivers to handle shader code, not your developer relations ability to optimize.
What is being consistently precluded: Futuremark could also have reason to reply in agreement with CM. Why could they? Because the shader output was identical. That's also why they couldn't validly express agreement with nVidia saying the same thing about their shader substitutions, and have any value as a comparative cross vendor benchmark.
My opinion: Their decision not to reply in agreement with CM makes 3dmark a better benchmark, but that's not because what CM is defending is "wrong" (and, in fact, seems "right" to me for actual games), but because I think general optimization ability is better in the long run than representing something that will vary, like devrel's efforts in games and/or game developer optimizing ability.
Again, that is a big distinction from what nVidia did, and JC and TS do directly agree with that. That CM is responding to the statements of a site that completely disregarded that distinction has everything to do with the validity of his statements. If he said the exact and full set of statements while addressing specifically Futuremark, I'd jump all over him. But...err...hold on a sec...he didn't. Ignoring where he said it, and to what he said it in reply, just ignores that simple fact.
I'd do that (after trying to ensure that it wasn't just a fragile general case optimization), because I'd agree with Futuremark's statement out of the two above. However, that doesn't mean that CM is clearly wrong, and TS and JC's lack of condemnation does lend validity to him arguing that. That's because they also exactly specified what is clearly wrong and directly established it as different than what ATI did.
"Morally grungy" is not "black and white" with regard to what can be defended or not. "Absolutely, positively wrong and indefensible" is. The latter was not applied to ATI, and by saying that nothing else matters except that both nVidia and ATI "cheated", that's exactly what you are ignoring about TS and JC's commentary.
You can't use "morally grungy" to propose that the opinion using that phrase supports the idea that something cannot be defended. Especially when the statement starts of with "acceptable" and goes to "range of defensibility" from that. That's pretty directly stating otherwise.
You can, however, use it to state that the statement using that phrase supports that it is wrong to say something can not be defended. If the "range of defensibility" sentence doesn't cut it for you, the observation that the full statement in question specifically said what can't be defended, and that it was something different, should.
This is very basic logic, and we have to be able to discuss what is right (or wrong) with the logic to hold a discussion.
I think Russ is doing the first, and CM is doing the second. If you can understand the first and second to which I refer, and then agree with the distinction between these two ideas, you can recognize that I've given my reasons for linking one to Russ and the other to CM. If you still disagree that this is valid, you can then go and address them and state why.
A reminder, since I can no longer find and reference a thread where I offered a similar explanation:
Simply repeating your assertion is not the same thing as addressing someone else's support for theirs. You don't have to agree with me, but it would be nice if you provided some addressable support if it so happens you don't.
Examples of support that is not usefully addressable: "your post is pointless", "your post discusses semantics" (assuming you don't state why semantics are actually bad in a particular case), "your post is long", "your post annoyed me", and "it is just my opinion, deal with it".
One explanation of the term "usefully addressable": where the validity of your opinion isn't dependent on the simple fact that it is you that holds it.
Which is why "you're repeating yourself", or other, ungarnished, directly observable facts, are usefully addressable when...er...they are directly observable independent of your evaluation.
My answer beforehand: Someone new started repeated what I was trying to address, and maybe discussing it on their terms will get me a useful reply.
Well, except that I addressed the assertion directly already earlier.
I addressed this earlier when he said it too. It is not a matter of getting it through to me, it is a matter of recognizing that I've already provided and supported an answer to it.
Gunhead said:
To which he could reply: that when the same optimizations would be done for games, and provide identical output, and in fact developer relations does exactly this for games, the label "inflating" is completely inaccurate.
To which Futuremark would likely reply: we want our benchmark to show the general ability of your drivers to handle shader code, not your developer relations ability to optimize.
What is being consistently precluded: Futuremark could also have reason to reply in agreement with CM. Why could they? Because the shader output was identical. That's also why they couldn't validly express agreement with nVidia saying the same thing about their shader substitutions, and have any value as a comparative cross vendor benchmark.
My opinion: Their decision not to reply in agreement with CM makes 3dmark a better benchmark, but that's not because what CM is defending is "wrong" (and, in fact, seems "right" to me for actual games), but because I think general optimization ability is better in the long run than representing something that will vary, like devrel's efforts in games and/or game developer optimizing ability.
Again, that is a big distinction from what nVidia did, and JC and TS do directly agree with that. That CM is responding to the statements of a site that completely disregarded that distinction has everything to do with the validity of his statements. If he said the exact and full set of statements while addressing specifically Futuremark, I'd jump all over him. But...err...hold on a sec...he didn't. Ignoring where he said it, and to what he said it in reply, just ignores that simple fact.
I'd do that (after trying to ensure that it wasn't just a fragile general case optimization), because I'd agree with Futuremark's statement out of the two above. However, that doesn't mean that CM is clearly wrong, and TS and JC's lack of condemnation does lend validity to him arguing that. That's because they also exactly specified what is clearly wrong and directly established it as different than what ATI did.
"Morally grungy" is not "black and white" with regard to what can be defended or not. "Absolutely, positively wrong and indefensible" is. The latter was not applied to ATI, and by saying that nothing else matters except that both nVidia and ATI "cheated", that's exactly what you are ignoring about TS and JC's commentary.
You can't use "morally grungy" to propose that the opinion using that phrase supports the idea that something cannot be defended. Especially when the statement starts of with "acceptable" and goes to "range of defensibility" from that. That's pretty directly stating otherwise.
You can, however, use it to state that the statement using that phrase supports that it is wrong to say something can not be defended. If the "range of defensibility" sentence doesn't cut it for you, the observation that the full statement in question specifically said what can't be defended, and that it was something different, should.
This is very basic logic, and we have to be able to discuss what is right (or wrong) with the logic to hold a discussion.
I think Russ is doing the first, and CM is doing the second. If you can understand the first and second to which I refer, and then agree with the distinction between these two ideas, you can recognize that I've given my reasons for linking one to Russ and the other to CM. If you still disagree that this is valid, you can then go and address them and state why.
A reminder, since I can no longer find and reference a thread where I offered a similar explanation:
Simply repeating your assertion is not the same thing as addressing someone else's support for theirs. You don't have to agree with me, but it would be nice if you provided some addressable support if it so happens you don't.
Examples of support that is not usefully addressable: "your post is pointless", "your post discusses semantics" (assuming you don't state why semantics are actually bad in a particular case), "your post is long", "your post annoyed me", and "it is just my opinion, deal with it".
One explanation of the term "usefully addressable": where the validity of your opinion isn't dependent on the simple fact that it is you that holds it.
Which is why "you're repeating yourself", or other, ungarnished, directly observable facts, are usefully addressable when...er...they are directly observable independent of your evaluation.
My answer beforehand: Someone new started repeated what I was trying to address, and maybe discussing it on their terms will get me a useful reply.
Well, except that I addressed the assertion directly already earlier.
I addressed this earlier when he said it too. It is not a matter of getting it through to me, it is a matter of recognizing that I've already provided and supported an answer to it.
