Pro/Forums - View Single Post

Cathar · 03-16-2006, 07:22 PM

I'd like to think that over IRC, whereby it's more conversational and people can get a better idea where each other is coming from, in the last two days we've managed to hash it out that we really all want the same thing at the end of the day. With the impersonal nature of internet forums, sometimes it's easy to lose track of where everyone wants to go, and once incendary comments start flying, all bets are off.

Sometimes it's just best to have a punch up, then when you're sick of beating your head or having it beaten, suck it up, say sorry, have a beer together (I guess IRC is the closest internet equivalent to having a beer together), and rediscover that you're all basically on the same side and this is achieving nothing.

Somewhere along the way I think everyone started second-guessing everyone else's "agenda", which is interesting, because there was only one person who was throwing that word around, and he no longer posts. I digress though. I don't want to play the blame game though, I'm as much as at fault for the shit-storm as anyone else.

I apologise to everyone for getting narky, and participating. I may have felt justified (and still do), but that doesn't excuse it.

At the end of the day, we're a community of people with like-minded interests. We all want to plot the best path forwards for exploring system-complete analysis, measurement, quantification, and improvements, just as we've been doing so very amicably for the last 4 years. We've had a disagreement over how that may best be achieved, and it all got far too personal when people started accusing others of having some "agenda", but again I'd like to think that we've worked past that misunderstanding.

Sure, some of us are still wearing bruises, and are still smarting, but I sense that we're on the upswing again, and that can only be a good thing for all.

Am looking forwards to new, mature, rational, reasoned, and non-attacking, threads that explore the technical issues that some of us have been discussing via IRC, and work to solve some of the technical frustrations that have been baffling forwards progress of late. There's room here for multiple independent points of view, and I believe that we may have reached a thresh-hold where there is no longer one clear answer for CPU waterblock performance assessment, but rather a set of co-related effects that can be split out and explored independently ("artificial but controlled carefully with set error-bars"), coupled with exploring them co-dependently ("real-world but with documented dependent effects").

No single method is more correct than the other, because what's correct depends on exactly what it is that is being quantified. Both serve separate goals that when brought together and studied carefully will provide a richer picture of the co-dependent effects of a real-world setup.

That's my 2c.

[Edit: Oh - and welcome back pH!]

03-16-2006, 07:22 PM	#12
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Re: OK I'm back I'd like to think that over IRC, whereby it's more conversational and people can get a better idea where each other is coming from, in the last two days we've managed to hash it out that we really all want the same thing at the end of the day. With the impersonal nature of internet forums, sometimes it's easy to lose track of where everyone wants to go, and once incendary comments start flying, all bets are off. Sometimes it's just best to have a punch up, then when you're sick of beating your head or having it beaten, suck it up, say sorry, have a beer together (I guess IRC is the closest internet equivalent to having a beer together), and rediscover that you're all basically on the same side and this is achieving nothing. Somewhere along the way I think everyone started second-guessing everyone else's "agenda", which is interesting, because there was only one person who was throwing that word around, and he no longer posts. I digress though. I don't want to play the blame game though, I'm as much as at fault for the shit-storm as anyone else. I apologise to everyone for getting narky, and participating. I may have felt justified (and still do), but that doesn't excuse it. At the end of the day, we're a community of people with like-minded interests. We all want to plot the best path forwards for exploring system-complete analysis, measurement, quantification, and improvements, just as we've been doing so very amicably for the last 4 years. We've had a disagreement over how that may best be achieved, and it all got far too personal when people started accusing others of having some "agenda", but again I'd like to think that we've worked past that misunderstanding. Sure, some of us are still wearing bruises, and are still smarting, but I sense that we're on the upswing again, and that can only be a good thing for all. Am looking forwards to new, mature, rational, reasoned, and non-attacking, threads that explore the technical issues that some of us have been discussing via IRC, and work to solve some of the technical frustrations that have been baffling forwards progress of late. There's room here for multiple independent points of view, and I believe that we may have reached a thresh-hold where there is no longer one clear answer for CPU waterblock performance assessment, but rather a set of co-related effects that can be split out and explored independently ("artificial but controlled carefully with set error-bars"), coupled with exploring them co-dependently ("real-world but with documented dependent effects"). No single method is more correct than the other, because what's correct depends on exactly what it is that is being quantified. Both serve separate goals that when brought together and studied carefully will provide a richer picture of the co-dependent effects of a real-world setup. That's my 2c. [Edit: Oh - and welcome back pH!]