PDA

View Full Version : Digit-life article using Unwinder's anti-detection script...TODAY!!!!


Pages : [1] 2 3

digitalwanderer
06-30-03, 10:07 AM
"All the kings horses and all the kings men...."

Today Digit-life is gonna go public with a VERY contraversial article according to a post at B3D (http://www.beyond3d.com/forum/viewtopic.php?p=136409#136409). It's been being discussed in a couple of threads on the board and Unwinder his ownself stopped in on a couple to comment.

Basically he's written a script to disable "optimization" detection in nVidia's drivers, and the Digit-life article is about the results of benchmarking 'em before/after/and against a 9800 Pro before/after.

I've been looking forward to it, methinks it'll be an interesting read....

(-boom!-) :eek:

Sazar
06-30-03, 10:11 AM
Originally posted by digitalwanderer
"All the kings horses and all the kings men...."

Today Digit-life is gonna go public with a VERY contraversial article according to a post at B3D (http://www.beyond3d.com/forum/viewtopic.php?p=136409#136409). It's been being discussed in a couple of threads on the board and Unwinder his ownself stopped in on a couple to comment.

Basically he's written a script to disable "optimization" detection in nVidia's drivers, and the Digit-life article is about the results of benchmarking 'em before/after/and against a 9800 Pro before/after.

I've been looking forward to it, methinks it'll be an interesting read....

(-boom!-) :eek:

it'll be interesting.. but the thing is.. is he turning off valid optimizations as well ? considering the way things are... every optimization is frowned upon... I just hope that those removed are explained in some detail instead of just saying... oh...it was application detecting so we got rid of it...

Behemoth
06-30-03, 10:15 AM
Originally posted by Sazar
it'll be interesting.. but the thing is.. is he turning off valid optimizations as well ? considering the way things are... every optimization is frowned upon... I just hope that those removed are explained in some detail instead of just saying... oh...it was application detecting so we got rid of it...
excellent point :D

The Baron
06-30-03, 10:16 AM
Yeah... I don't think all application-specific optimizations are evil. Here are the EVIL cases:

1. if it reduces quality
2. if it's specifically for a benchmark

Anything else is kinda meh. I mean, if they write something that makes UT2k3 run 15% faster regardless of the situation, and it has no bearing on quality, WHY WOULD YOU COMPLAIN?!

Plus, I have an odd feeling that a lot of them are either workstation-class software or for compatibility issues...

Behemoth
06-30-03, 10:30 AM
Originally posted by The Baron
Yeah... I don't think all application-specific optimizations are evil. Here are the EVIL cases:

1. if it reduces quality
2. if it's specifically for a benchmark

Anything else is kinda meh. I mean, if they write something that makes UT2k3 run 15% faster regardless of the situation, and it has no bearing on quality, WHY WOULD YOU COMPLAIN?!

Plus, I have an odd feeling that a lot of them are either workstation-class software or for compatibility issues...
aint 27 shaders/detections of them belong to shadermark already? :)

digitalwanderer
06-30-03, 10:33 AM
Originally posted by Sazar
it'll be interesting.. but the thing is.. is he turning off valid optimizations as well ? considering the way things are... every optimization is frowned upon... I just hope that those removed are explained in some detail instead of just saying... oh...it was application detecting so we got rid of it...
I have no clue, I haven't seen any of it yet either....but you raise some valid points and I hope they explain in excruciating detail EXACTLY what his anti-detect script does and how, because from the threads I HAVE seen about it ATi is gonna get smacked pretty hard to by it. :(

I'm just as interested as everyone else about seeing what the what is with this, and just as nervous/skeptic right now...but I thought it was definately worth mentioning it since I came across that post since it is gonna be interesting. :cool:

Sazar
06-30-03, 10:39 AM
Originally posted by digitalwanderer
I have no clue, I haven't seen any of it yet either....but you raise some valid points and I hope they explain in excruciating detail EXACTLY what his anti-detect script does and how, because from the threads I HAVE seen about it ATi is gonna get smacked pretty hard to by it. :(

I'm just as interested as everyone else about seeing what the what is with this, and just as nervous/skeptic right now...but I thought it was definately worth mentioning it since I came across that post since it is gonna be interesting. :cool:

its gonna be really funny to see ati and nvidia agreeing on something if they both get smacked down by this :D and saying unwinder's anti-optimization thinger is fubar... lol...

/me waits patiently...

digitalwanderer
06-30-03, 10:54 AM
Originally posted by Sazar
its gonna be really funny to see ati and nvidia agreeing on something if they both get smacked down by this :D and saying unwinder's anti-optimization thinger is fubar... lol...

/me waits patiently...
I don't know, I got an awful lot of respect for Unwinder...the man freaking KNOWS drivers! :eek:

/me waits impatiently

/me bites me nails

Hellbinder
06-30-03, 11:50 AM
I think the whole thing is just stupid. There is nothing wrong with application detection. I even Question wether lowering IQ for speed is wrong.

As long as Reviewers are *honest* and say.. "Look this card is faster becuase it looks like ass". Then what is the problem. Yes overall The most desireable end is clean pure code with perfect support for all games without tweaking the games appearance etc at all.

At this point i think it is getting really really RIDICULOUS. I for one DONT CARE if the Nv3x cards are slower without Application detection. The people that Buy the cards dont care either. If the game looks good and is fast thats all that matters.

Sazar
06-30-03, 11:53 AM
Originally posted by Hellbinder
I think the whole thing is just stupid. There is nothing wrong with application detection. I even Question wether lowering IQ for speed is wrong.

As long as Reviewers are *honest* and say.. "Look this card is faster becuase it looks like ass". Thgen what is the problem. Yes overall The most desireable end is clean pure code with perfect support for all games without tweaking the games appearance etc at all.

At this point i think it is getting really really RIDICULOUS. I for one DONT CARE if the Nv3x cards are slower without Application detection. The people that Buy the cards dont care either. If the game looks good and is fast thats all that matters.

hmm.. refresh my memory... but werent you the one who started multiple threads in forums such as this/rage3d/b3d saying how evil ati was for having such and such optimizations in cat 3.4 and also said quite a bit about how you thought application detection was so evil ?

frankly speaking m8... though I agree with some of the points you put out... you go from one extreme to the other extreme so fast it becomes quite difficult to keep track of what exactly it is that you are saying :)

Nutty
06-30-03, 11:56 AM
wow! When I saw your name there HellBinder, I thought you were gonna start laying into nv about this whole thing.

I agree, we're soo used to seeing cpu benchmarks. CPU's _have_ to do it perfect, or apps will simply crash, and not work.

Having pixels that are slightly off from perfect doesn't cause any problems at all.. You gotta weigh up the balance between pixel quality and speed.

If an IHV can redo a games shader to make it run better for their customers with negligable IQ degradation, is it soo bad that they do it? I dont think it is.

jbirney
06-30-03, 11:56 AM
Sazar


hellbinders post over at Rage were about application detection for Benchmarks. App dec for games fine. App dec for benchmarks very evil.

Sazar
06-30-03, 12:02 PM
Originally posted by jbirney
Sazar


hellbinders post over at Rage were about application detection for Benchmarks. App dec for games fine. App dec for benchmarks very evil.

I understand the point jbirney :)

like I said.. I agree with some of the points he put out there... I have no problem with his logic... I question the way he puts them out there... a couple of weeks ago he was singing a different tune or so it seemed :)

perhaps if the posts were toned down and put forth more as opinions rather than facts it would be a lot easier to just overlook this...

frankly I am of the opinion that you have put forth... and I have maintained that line of thinking for a while... :)

I'll leave it @ that.. didn't mean to make my post sound like a flame...

Hellbinder
06-30-03, 12:10 PM
hmm.. refresh my memory... but werent you the one who started multiple threads in forums such as this/rage3d/b3d saying how evil ati was for having such and such optimizations in cat 3.4 and also said quite a bit about how you thought application detection was so evil ?

frankly speaking m8... though I agree with some of the points you put out... you go from one extreme to the other extreme so fast it becomes quite difficult to keep track of what exactly it is that you are saying

You got the Wrong guy and the wrong message....

I was upset at Cat 3.5 because i beta tested them for 4 weeks and was never told anything about GLexcess tweaks for 80% increases in speed. When i beta test its on GAMES not benchmark programs. Which is why people generally didnt understand why i freaked out about it. Here we are hammering Nvidia day after day. Talking about Tweaking for benchmark apps and what a waste of time it is.. and how is sort of evil... And the latest ATi drivers ave an 80% increase in 2 benchmarks. With everything else pretty much the same.

reever2
06-30-03, 12:12 PM
Originally posted by Sazar
it'll be interesting.. but the thing is.. is he turning off valid optimizations as well ? considering the way things are... every optimization is frowned upon... I just hope that those removed are explained in some detail instead of just saying... oh...it was application detecting so we got rid of it...

I guess thats where image quality and extreme zooming come in, but knowing its digit-life they would probably skip all that stuff

Hellbinder
06-30-03, 12:13 PM
I question the way he puts them out there... a couple of weeks ago he was singing a different tune or so it seemed

Again no.. Go back and read any post of mine you want from ANYWHERE. I ALWAYS have said that application detection is GOOD. It only Turns bad depending on what you do with it.

I swear.. If more of you paid attention to my actual content instead of just assuming i am a mindless fanboy you would maybe learn something about what actually motivates me.

Ratchet
06-30-03, 12:28 PM
My basic views:
App detection for game = good (as long as you don't reduce image quality. Let the use do that if her decides he'd rather have performance than image quality)
App detection for benchmarks = bad. No if's and's or but's about it. If you're detecting a benchmark app, you should be called out on it and labeled a cheater.

There's also the very very lame idea of detecting timedemos in games that are used for benchmarks. This goes beyond the lame underhanded fraud of synthetic benchmark app detection and should get the offending bastards a good swift kick in the nuts for every day that they've lied to us.

digitalwanderer
06-30-03, 12:38 PM
Originally posted by Ratchet
My basic views:
App detection for game = good (as long as you don't reduce image quality. Let the use do that if her decides he'd rather have performance than image quality)
App detection for benchmarks = bad. No if's and's or but's about it. If you're detecting a benchmark app, you should be called out on it and labeled a cheater.

There's also the very very lame idea of detecting timedemos in games that are used for benchmarks. This goes beyond the lame underhanded fraud of synthetic benchmark app detection and should get the offending bastards a good swift kick in the nuts for every day that they've lied to us.
A very sensible and reasonable sounding approach to me, the Dig officially approves of the Cheater Nut-Kicking solution! :cool:

reever2
06-30-03, 12:39 PM
Originally posted by Ratchet

There's also the very very lame idea of detecting timedemos in games that are used for benchmarks. This goes beyond the lame underhanded fraud of synthetic benchmark app detection and should get the offending bastards a good swift kick in the nuts for every day that they've lied to us.

Hrm I guess I forgot about that too. what Digit-life needs to do is use custom timedemos along with the regular timedemos and use the anti-detection script for both of them, and if the custom timedemos dont get any performance decrease and the regular ones do, then that would be cause for concern

dan2097
06-30-03, 01:03 PM
I really hope this article is good but I bet it wont be.

Firstly they must try and disable app specific optimizations for both Nvidia AND ATI.

Secondly disablement of application specific optimizations for changes in performance should only be done on benchmarks.

When looking at actual games all that is important is that image quality does not improve noticeably when the optimizations are disabled.

Investigating whether the common game time demos take a larger performance hit than custom time demos when app specific optimizations are disabled would be interesting though.

digitalwanderer
06-30-03, 01:04 PM
Even if the article totally blows it'll at least give us all something to argue about for the day. :cool:

prodikal
06-30-03, 01:53 PM
ROFL Fighers to your corners lol

poursoul
06-30-03, 04:17 PM
Originally posted by digitalwanderer
Even if the article totally blows it'll at least give us all something to argue about for the day. :cool:

we could always just argue about this (http://www.hardocp.com/).

GlowStick
06-30-03, 04:39 PM
Cant wait to see this released, we finaly get to see in depth tests (hopefully)

Sazar
06-30-03, 05:33 PM
Originally posted by Hellbinder
Again no.. Go back and read any post of mine you want from ANYWHERE. I ALWAYS have said that application detection is GOOD. It only Turns bad depending on what you do with it.

I swear.. If more of you paid attention to my actual content instead of just assuming i am a mindless fanboy you would maybe learn something about what actually motivates me.

I don't consider you to be a mindless fanboy m8... if you actually go back through all the threads you have ever posted... I am sure you will notice I have stuck up for you on more than one occasion :)

if you posted in a clear manner without all teh caps ( == shouting) it would make it a lot easier to understand some of your posts in the manner intended... a number of posts put out there with your opinions seem to be posted as fact.. which makes it that much harder to form an opinion on my own :) ...

like I said... I did not mean for my post to sound like flaming. and I stand by that... I dont agree with everything you say... and I don't disagree with everything you say...

you have explained your end... and thats all I really could have asked for... a clarification of the point you made... thanks to jbirney as well for pointing out some things...

I am still learning about gpu's and what not.. it is an interesting field and I rather enjoy having discussions or just reading the discussions that take place... to me that is the best learning experience...

now... I am just going to wait for that article :)