Dave H,
I'm afraid I categorically disagree with some fundamental statements of your post.
...
IMO a whole lot of the attitude Russ objects to comes from an ignorance of the challenges and time-frames involved in designing and manufacturing a high-performance ASIC as complicated as a GPU.
...
It was indeed the basis of the commentary Russ provided that ignorance of reality is necessary for condemnation. This is what I disagreed with, especially as simplification was the means for this relationship being proposed.
nVidia's situation and lack of competitiveness is unique, and so is the magnitude of measures they've taken in response to it. Yes, it is realistic to view this as expected behavior, but it is not realistic to presume this is the only thing a company is capable of (as demonstrated by other companies acting differently). Even if you propose this as the only thing that could be expected of this particular company in their particular situation, this doesn't do anything to address being strongly disgusted by a company whose decisions put them in this situation (the aggressive commitment to PR was a facet of their success as well), nor necessitate that anyone who holds that opinion be automatically "ignorant of reality" as this discussion continues to propose.
And you don't even successfully propose that maintaining obscurity was their only alternative to what they've done, as you propose as the reason for them "having no other option", as there are plenty of marketing efforts that could have been more successful at this point than what they did do, and this has been evident since at least last year November. The lack of the ability of management and marketing to realize the insufficiency of their hardware for what they did try to do is plenty of reason for condemnation, by consumers and stockholders alike.
The problem with the objections Russ puts forth is that he doesn't just respond with pointing out realism, he dictates that there is no foundation for condemnation in the face of realistic evaluation, by the expedient of simplifying reality to say any companies reacting to their situation is the same as nVidia reacting to their situation, because "reacting to the situation" can be used for both. Nevermind the company policies that put them there, or the possibility of alternative to their approach and any evaluation of how evident alternatives were. Pointing out realism would have stopped at pointing out the error, and please note the distinction already made in this regard in my prior responses.
As for your own commentary...as far as it goes in pointing out the error in blithe assumptions concerning how the NV3x can have a magic wand waved at it and be fixed, I agree. Where you perceive the validity in going further and stating that this reality precludes vehement contempt doesn't hold together and contradicts some of your own statements, so I must therefore ask if I misunderstood you in some regard in your complete agreement with Russ, and your discussing only ignorance as a factor in the attitude that Russ responded to?
Actually, I also disagree with your premise of uncompetitive situation, at least prior to their self-destruction of consumer confidence in their brand name. Much of their technical inferiority is well within the ability of loyalists and ignorant consumers to overlook, and it is only their aggressiveness in unlimited abuse of ignorance that resulted in its erosion. Running things like Quake III benchmarks and winning while spinning things like the Splinter Cell shadow implementation issues would have been truthful things with plenty of mileage for addressing your concerns. A stance on 3dmark that stopped before the aggressive lying, but still put forth that nVidia found it unrealistc and depended on factual presentations about why "it didn't represent gaming" by a focus on titles at the time would have met much the same short term success, but without the backlash that seems to have destroyed the benefits to them of the actions they did take.
And I don't mean by more successful deception, but concentrating more effort on true advantages (which isn't restricted to the items you list for their product line up).
Their situation in HL 2 and Doom 3's delay (assuming Doom 3's delay has nothing to do with their negative actions) would likely be similar, but the perception of the company at the time of that situation could have a completely different complexion than it does now...they didn't have to damage their positive brand perception as thoroughly as they have, and options to protect it more than they have would have resulted in less condemnation from everyone (or, in the case of anti-nVidia sentiment that doesn't limit itself by their actions, less opportunity for condemnation to be validated), and far more opportunity for distasteful "journalism" to work for them successfully (which would have kept me as thoroughly out of their consumer group, but not a large portion of enthusiasts).
This was an option for the company, as well, and taking the easily chosen course of committment to unlimited deception based on the premise of consumer ignorance does not seem to be a virtue for nVidia, their stockholders, or costumers. I'm not sure who it can be said should
not be disgusted with nVidia, which is why I categorically disagree with some of what you seem to propose.
Their response to their situation is a decision on their part, and the situation itself is a failure on their part. Their situation is the result of their company decision-making and policies, and proposing their situation as a given that dictates their response doesn't seem pertinent at all to condemning the decisions that put them in it, even ignoring the possibility of alternatives. Perhaps there is something wrong with a company that makes such decisions? What about people who condemn because of that...why does your commentary presume their ignorance?
I recommend that if your intent is to correct people making erroneous assumptions about the NV3x being easily re-engineered, you restrict commentary to their error in analyzing nVidia's situation instead of proposing that condemnation of nVidia is necessarily precluded or ameliorated by that situation. They are not the same thing. I recommend the same to Russ.
As you should know if you are familiar with my commentary throughout this affair, I expressed my confidence in their ability to achieve at least the 256-bit bus with the NV35 in a timely fashion, and was impressed with the (minor, but significant) improvement they did achieve beyond a 256-bit bus in the given time. I also continue to propose the possibility of dividends in some of their decisions in a new design such as the NV40, and I even thought some degree of register issue or render target solution is conceivable in the NV38 (until the silence surrounding it persisted, though some sort of coordination with Det 50's to associate improvements with the driver might be possible still).
Yet I still condemn them, and I'm left wondering if my expectations of improvements is being proposed as ignorance by you by the nature of your reply, without an intervening discussion of the proposal itself. This is a problem (IMO) with depending on generalizations as the foundation of an argument, and characterizing someone else's suppositions and ignorance without query as a premise for it.
If your post didn't seem to propose complete agreement with the statements Russ made and that maintaining vehement condemnation of nVidia required ignorance of reality, I wouldn't be replying to you as I did above.
Whether you agree or disagree, I recommend you address the reasons any person provides specifically, instead of proposing an argument over the reasons for nVidia doing what they did as a substitute for evaluating why an emphatically negative reaction to what they did might be justifed, even without "ignorance" of the reasons. Proposing the implicit relationship skips over too much.
