Minor point. The GBA version gets a steady fps rating, whereas the N-gage version is jerky as hell (I won't pretend to know what the exact numbers are, but when things get busy the N-Gage version is dangerously close to unplayable)
I don't think you are living in the real world here...
Yes we don't like AOL and other large ISPs, but if your little ISP can't send e-mail to AOL users then to be honest that is a large segment of the internet. I know lots of people on AOL, as I'm sure most people do. If I couldn't e-mail them, or recieve e-mail from them I'd have to change ISP, simple as that.
Is there a super optimized compiler for the PPC-970? If so, I'd be more than happy to see some benchmarks from it.
It is my belief that gcc is really not that good on x86, espically as you get up to the pentium 4, because it is such a complex chip. The PPCs have a simpler architecture (which yes, is a good thing) so I suspect gcc does better there.
OK, I got my results by compairing their results to the results I get using icc on a Dual P4 3.2Ghz.
I only ran SPECfp_base2000 (because it was the only test I had around)
I got 16.1 (higher than the mac result at 15.7, from the paper)
Argueably I was unfair for using icc instead of gcc, but icc is so much better, why should I use an inferior compiler? If there is a better compiler for mac then I'd be happy to see the results.
I don't agree that using gcc on both platforms somehow makes things "fairer" as it will clearly have to optimise for the different processors differently, so why not use the best compiler you can?
Remember those "We have the fastest machine ever" benchmarks (that were a bit dodgy anyway..), well a quick look now shows that Intel have since then continued ramping up the old processors and these G5's aren't actually the fastest machines now you can actually buy them. Oh well.
Having works in a games company, I can at least tell you what may opinions of why that wouldn't work are. Most games do start with a nice seperated interface, but the it gets destroyed for 3 main reasons.
1) Performance - you want to squeeze every last bit out of the latest shiny card that has come out and you suspect the reviewers will be using.
2) Driver problems - similarily, you have to work around the problems that there are in most directX drivers (and don't suggest doing openGL only, because the openGL drivers are often worse!)
3) Deadlines - in that last month, it's a case of slap in anything to make it work.. and if it's a choice between slipping a couple of weeks or dumping linux (which it almost certainly will be), the dumping linux it will be
4) connected to 3, with windows "all" you have to do is check windows 95,98,ME,2000,XP (which is still a lot!) With linux there isn't really a limit to the number of versions you could have to test on...
Re:Goedel says benchmarks are inherently flawed.
on
Examining Benchmarking
·
· Score: 4, Insightful
Seriously? Who modded this up?
Godel's incompleteness theorem is about complex mathematical systems and the essense of proof. You don't need an external object to compare to, anything will do. You just choose some graphics card as your fixed point and then compare everything to that card.
I would say that console games are much better. The developer is working for one controller only, so can set up the controls to be much more usable. For example see Zelda: Wind Waker, where almost the entire game is played through 5 buttons.. I'm sure on the PC it would spread over half the keyboard!
think this is exactly the point: have a computer with lots of ready-to-use-software, OS, libraries, and you don't learn nearly as much as if you need to write all those nifty things yourself. And let a beginner use somethink like OpenGL/DirectX8 and they won't understand simple basics like "How do I draw a 3D cube on a 2D display?"
While I realise this is somewhat anti-geek, is this really nessasary any more? A few years ago, everyone had to learn assembler. Now I feel there is no need (other than interest) to actually know any assembler at all.
Also of course how you draw a 3d cube on a 2d display in (very old) software and how it is done in hardware are so different I don't think you would really learn much anyway.
The error is in thinking that a computer should be intuitive. Computers are equivalent to Turing machines, modulo the bounded memory; they can go far, far beyond our intuitions. The only way to make them intuitive is to dumb them down, i.e. limit what they can do. So be prepared to choose between having your computer dumbed down to a consumer appliance or else having to learn a lot in order to master it.
Why shouldn't computers be intuative? Obviously I don't expect writing a full program to be simple, but it is my belief that writers of software should, within reason, try to do as much as possible to make computers as easy to use as possible.
So... they're demanding to sue en masse like this? Using lawsuits and demanding massive settlements? Isn't this the definition of barratry---abuse of the legal system for extortion? If so, do smartcard reader owners have the basis for a class action?
No, because if you'd read the article (and related articles) you would know they are only sueing people who bought equipment which was specifically advertised as having the use of cracking DirecTV's system. If you have a legal use for such a product you shouldn't buy it from someone who is specifically advertising it as being illegal.
Hmm... I wonder. Could the slowdown have been due to the famous "slashdot effect", and in fact have nothing to do with your computer, or their website design (except the fact they can't handle a slashdoting, which few can.. in fact they are handling it better than many)
If you don't want your work to be copied, you really have only two choices. Either don't disclose it without strict NDAs, contracts, etc. (i.e., make it a trade secret). Or do work that does not consist of a very large and easily copyable integer.
The problem is making good versions of those "large integers" requires a lot of skill, dedication and time, often with huge teams of people.
The days of profit owing solely to artificial scarcity are over, and no amount of wishing or legislating to the contrary is going to change that.
Surely the advantage of "large integers" is that there isn't a scarcity, I can always make another copy. If what you mean is "I think I deserve all software and music for free, because I find that copying it doesn't take much effort" then I think you have to think a bit more about what you are saying...
Just because a crime is easy to get away with doesn't make it any more right. I've just finished playing Zelda:The Wind Waker. I thought it was an excellant game. I'm sure it took a large number of people a long time to write. In the future everything is going to be "a big integer", so if anything it is more important we get the ideas of how to protect them down correctly. We don't want to go overboard, but we have to protect people who work hard on these things too.
Not really a lot I can say, but seeing as I seem to be first(ish) here, I'll say some stuff anyway:)
It is nice to see this kind of coverage in the popular media. It no longer seems as ridiculas as it did to imagine a general shake-up of patents, which is good. The article also describes the problem well, with good examples (as opposed to some of the more usually used stupid examples).
I don't think we should get rid of patents, I don't even think "software patents" are a bad thing (if anyone tries quoting the 'all software is produced by a turing machine, so is all obvious argument, I'll hit them!), but hopefully we can reach a sensible system!
There is nothing stopping linux getting e-mail trogans. There is nothing stopping such a program emailing lots of people from your pine / mozilla / whatever address book, then deleting all the files in your user space. For single users linux isn't really any better than windows from a virus point of view...
Yes, but by an identical proof you can prove that any piece of text that is ever written is just as "obvious", and this will clearly include every patent:)
Nice try, icc has compiled 2.4 for some time, and will compile 2.6 I expect. The only reason it doesn't compile 2.5 is because it is such a moving target and its hard to tell if it breaks if it is due to an intel, GCC or kernel bug.
Slashdot Mirror
Minor point. The GBA version gets a steady fps rating, whereas the N-gage version is jerky as hell (I won't pretend to know what the exact numbers are, but when things get busy the N-Gage version is dangerously close to unplayable)
I don't think you are living in the real world here... Yes we don't like AOL and other large ISPs, but if your little ISP can't send e-mail to AOL users then to be honest that is a large segment of the internet. I know lots of people on AOL, as I'm sure most people do. If I couldn't e-mail them, or recieve e-mail from them I'd have to change ISP, simple as that.
Is there a super optimized compiler for the PPC-970? If so, I'd be more than happy to see some benchmarks from it. It is my belief that gcc is really not that good on x86, espically as you get up to the pentium 4, because it is such a complex chip. The PPCs have a simpler architecture (which yes, is a good thing) so I suspect gcc does better there.
OK, I got my results by compairing their results to the results I get using icc on a Dual P4 3.2Ghz.
I only ran SPECfp_base2000 (because it was the only test I had around)
I got 16.1 (higher than the mac result at 15.7, from the paper)
Argueably I was unfair for using icc instead of gcc, but icc is so much better, why should I use an inferior compiler? If there is a better compiler for mac then I'd be happy to see the results.
I don't agree that using gcc on both platforms somehow makes things "fairer" as it will clearly have to optimise for the different processors differently, so why not use the best compiler you can?
MrJeff
Remember those "We have the fastest machine ever" benchmarks (that were a bit dodgy anyway..), well a quick look now shows that Intel have since then continued ramping up the old processors and these G5's aren't actually the fastest machines now you can actually buy them. Oh well.
Are you sure that the X-box / PS2 / Gamecube will fair any better? Having a quick look through a list of games I'm not convinced they would.
Having works in a games company, I can at least tell you what may opinions of why that wouldn't work are. Most games do start with a nice seperated interface, but the it gets destroyed for 3 main reasons.
1) Performance - you want to squeeze every last bit out of the latest shiny card that has come out and you suspect the reviewers will be using.
2) Driver problems - similarily, you have to work around the problems that there are in most directX drivers (and don't suggest doing openGL only, because the openGL drivers are often worse!)
3) Deadlines - in that last month, it's a case of slap in anything to make it work.. and if it's a choice between slipping a couple of weeks or dumping linux (which it almost certainly will be), the dumping linux it will be
4) connected to 3, with windows "all" you have to do is check windows 95,98,ME,2000,XP (which is still a lot!) With linux there isn't really a limit to the number of versions you could have to test on...
Seriously? Who modded this up?
Godel's incompleteness theorem is about complex mathematical systems and the essense of proof. You don't need an external object to compare to, anything will do. You just choose some graphics card as your fixed point and then compare everything to that card.
I would say that console games are much better. The developer is working for one controller only, so can set up the controls to be much more usable. For example see Zelda: Wind Waker, where almost the entire game is played through 5 buttons.. I'm sure on the PC it would spread over half the keyboard!
While I realise this is somewhat anti-geek, is this really nessasary any more? A few years ago, everyone had to learn assembler. Now I feel there is no need (other than interest) to actually know any assembler at all.
Also of course how you draw a 3d cube on a 2d display in (very old) software and how it is done in hardware are so different I don't think you would really learn much anyway.
Why shouldn't computers be intuative? Obviously I don't expect writing a full program to be simple, but it is my belief that writers of software should, within reason, try to do as much as possible to make computers as easy to use as possible.
No, because if you'd read the article (and related articles) you would know they are only sueing people who bought equipment which was specifically advertised as having the use of cracking DirecTV's system. If you have a legal use for such a product you shouldn't buy it from someone who is specifically advertising it as being illegal.
Hmm... I wonder. Could the slowdown have been due to the famous "slashdot effect", and in fact have nothing to do with your computer, or their website design (except the fact they can't handle a slashdoting, which few can.. in fact they are handling it better than many)
Damn you!
*thwack*
The problem is making good versions of those "large integers" requires a lot of skill, dedication and time, often with huge teams of people.
The days of profit owing solely to artificial scarcity are over, and no amount of wishing or legislating to the contrary is going to change that.
Surely the advantage of "large integers" is that there isn't a scarcity, I can always make another copy. If what you mean is "I think I deserve all software and music for free, because I find that copying it doesn't take much effort" then I think you have to think a bit more about what you are saying...
Just because a crime is easy to get away with doesn't make it any more right. I've just finished playing Zelda:The Wind Waker. I thought it was an excellant game. I'm sure it took a large number of people a long time to write. In the future everything is going to be "a big integer", so if anything it is more important we get the ideas of how to protect them down correctly. We don't want to go overboard, but we have to protect people who work hard on these things too.
whitehouse.gov
Not really a lot I can say, but seeing as I seem to be first(ish) here, I'll say some stuff anyway :)
It is nice to see this kind of coverage in the popular media. It no longer seems as ridiculas as it did to imagine a general shake-up of patents, which is good. The article also describes the problem well, with good examples (as opposed to some of the more usually used stupid examples).
I don't think we should get rid of patents, I don't even think "software patents" are a bad thing (if anyone tries quoting the 'all software is produced by a turing machine, so is all obvious argument, I'll hit them!), but hopefully we can reach a sensible system!
There is nothing stopping linux getting e-mail trogans. There is nothing stopping such a program emailing lots of people from your pine / mozilla / whatever address book, then deleting all the files in your user space. For single users linux isn't really any better than windows from a virus point of view...
Yes, but by an identical proof you can prove that any piece of text that is ever written is just as "obvious", and this will clearly include every patent :)
and here we see your plan's tragic flaw... good try tho
Nope, with the subscription system I've seen sites go down before the story even goes on the front page :)
Although of course del * is all you need as a regular expression.
Actually, this brings a thought to mind..
If SCO own all this unix stuff, I assume they also own all the common unix tools?
And as they seem to be dragging in everything they can, what's the chances they'll drag in GNU's grep / ls / cat / etc...?
Just a thought before RMS starts saying he doesn't care about linux getting dragged down by SCO.
Nice try, icc has compiled 2.4 for some time, and will compile 2.6 I expect. The only reason it doesn't compile 2.5 is because it is such a moving target and its hard to tell if it breaks if it is due to an intel, GCC or kernel bug.
You, know I reallly wish there was a:
-5 : They asked me to
Moderation option...