The doomsayers were right. USENET is a vast wasteland now. Ask your mommy and daddy what it used to be like before the Spammers destroyed it.
Maybe I'm just lucky with the 3 newsgroups I read, but none of them have more than maybe 1% spam. Literally: a couple of spams a week. They're not moderated newsgroups or anything, either - just one aus.* local group, one in the rec.games.* heirarchy, and one in comp.lang.*
The reports of Usenet's death are greatly exaggerated, if you ask me. I read news with Mozilla, so I don't even have a killfile! And yet, I'm perfectly happy.
Charged at $8 per gigabyte over the cap. Reliable estimates by people in the know estimate that Bell's actual costs for bandwidth are in the range of 50 cents to $1 per gigabyte.
Sucks to be you. Not. My Telstra ADSL account, touted as the model these American ISPs are talking about emulating, charges 13.9c per MEGAbyte over the cap - that works out to about $A142 per gig.
So you can understand that I'm very careful not to go over the limit. Of course, because I hate Telstra, come the last day of the month I'm downloading any old crap just to make sure I get as close to the limit as I can. Don't want to let them off providing less bandwidth than I'm paying for!
Seriously, if I could get a 19" LCD for $200-300 more than a 19" CRT with the same resolution, I'd do it.
Bear in mind that a 19" LCD will have a larger viewable area than a 19" CRT. Try comparing a 17" LCD to a 19" CRT, you'll probably find that they look much the same size.
I went from a 17" CRT to a 15" LCD (Hercules Prophetview), and it looks to be much the same size, and a hell of a lot easier on the eyes. It only does 1024x768 natively, but that's good enough for me, and running games at lower res actually looks excellent - the panel bilinearly filters it up to 1024x768 (no crappy pixel doubling), and I think 640x480 for instance looks better than it did on a CRT that could display it "properly".
Hm, nice try, but here [vuurwerk.net] we can see that the N64 uses a MIPS R4300i, which is a 64-bit device.
I'm sorry, I neglected to mention that I actually live in a parallel universe where the current crop of Nintendo game consoles is made up of machines called "Gamecubes", and where "N64s" are actually the previous crop of Nintendo game consoles.
In my world, these "Gamecubes" use a CPU called Gekko, which is an IBM-developed PowerPC family chip, with 32-bit registers.
Aren't the current crop of Nintendo game consoles powered by a 64-bit MIPS?
No, they are powered by a 32-bit PowerPC-family chip.
Most variables have a minimum of 8 significant bits. The average length of a character string is in the 8-12 byte range (64-96 bits!)
..which almost always need to be operated on one character, i.e. 8 bits, at a time.
Third, what's this crap about it being "too expensive" to transfer 64-bits of data in from RAM? All modern processors have 64-bit wide data busses
..and said transfers are used to fill cache lines, not to read individual variables. With 64-bit variables, you fit half as many into each of those cache lines, and thus go crashing back to memory accesses twice as often.
The way I see it, Divx needs 3 things before it becomes a major threat to DVD.
1-Players capable of playing multiple soundtracks, for multiple languages and/or commentary.
Well, I don't know about 2 and 3, but you can do that one now - you just need to put the DivX video into an OGM (Ogg Multimedia) container rather than an AVI container. Then you can have multiple soundtracks (and they can be VBR audio too, whereas AVI only works properly with CBR).
That guy? I reckon this guy owes us an explanation!
Re:32,504 800 MHz G4 vs. 45,998 2 GHz Athlon XP?
on
RC5-64 Success
·
· Score: 2
800 MHz G4 is faster crunching the keys than a 2 GHz Athlon XP. I am reading that right?
Yes. I've never seen anything which shows off Altivec quite as well as RC5 cracking. There are hand optimised assembly cores for various CPUs in the d.net client, but the Altivec-enhanced G4 core pretty much destroys everything. I expect it's because Altivec has vastly more flexible shuffling instructions than MMX.
This by no means proves that a G4 is "better" than an Athlon, but it's interesting.
Yep, you're retarded.:-) Nah, seriously, the pixel fill rate is 2600 (million pixels/sec) compared to 1100, 1200 and 880 for the three compared cards. So, yes, more than double that of the closest competition.
You're looking at the texel fillrate, which is also 2600 (million texels/sec) for the Radeon 9700, and is not blistering ahead of the competition (indeed, as you say, it's below one of them, the Matrox Parhelia).
But, as they say a few paragraphs down, it's better to get a high max texel fill rate from high clock speed and lots of pipes (as the Radeon 9700 does) rather than from lots of texture units per pipe (as the Parhelia does), because not every game is going to want to use 4 textures on every pixel. They all want to draw a helluvalot of pixels, though.
On a my P3-800 with 256mb of ram and a geforce 2 ti it ran like ass anytime there were more than 3 guys running around or if I was in a big room. Really pissed me off. And that was at 800x600, I had to turn it down from 1024x768 because it was unplayable.
News flash: if changing resolution improves performance, then your problem is that you're fillrate bound on the graphics card. Nothing to do with your CPU, nothing to do with "lazy/inefficient programming".
If you were getting the same crappy performance regardless of resolution, then you'd have a point.
"Why fix what ain't broken" is a damn good way to sum it up, IMO. This is coming from a guy who's perfectly happy running MacOS 8.6.1 on his G4, and WinME on his Windows boxes.
Many would say that you broke your Windows boxes when you "upgraded" to WinME from the far superior Win98.
While I applaud the open source community for producing such a high-quality competitor to MP3 as OGG, the real issue of getting people to switch still lies in hardware support and easy-to-use, CDDB compatible OGG CD-rip utilities.
Hardware support is the biggie. CD ripping is a non-issue - Exact Audio Copier and CDex, which are surely the two rippers most people are using, both support Ogg perfectly well (along with a swag of other non-MP3 formats).
I hate to question the logic of engineers, but it does seem pointless to me to keep the cache size the same, but make it 8-way set associative.
Well, I guess the important thing is that having the 128KB cache instead of 256KB saves money (obviously vital for a games console), whereas going with the 4-way cache instead of 8-way wouldn't save much if anything.
However, you do sometimes want to cache two address with the same lower bits, so to do this they came up with the idea of 2-way set associative, where you divide the memory in half, and make half of it a place to put the second-address with the same lower bits.
2-way set associative means that each piece of memory can be cached in two different places (and when it needs to be cached, some algorithm e.g. Least Recently Used will be used to decide which of those places it should be cached). This is unquestionably a good thing, it means you can avoid ping-ponging if two subroutines being called alternately both happen to map to the same cacheline. It certainly doesn't mean that you can only cache half as much stuff or anything bad like that.
8-way set associative is quite high.. probably high enough that as long as your code fits in cache, you won't get any evictions due to different chunks of code fighting over the same cachelines..
There can be a performance drawback to increasing n in an n-way set associative cache, as the increased complexity of the cache lookup logic means that it can be slower to retrieve a cacheline. I have no idea if there is a difference in practice between the P-III and Celeron L2 caches, though.
Pentium III ? More likely Celery^Hon according to this screenshot...
Actually the Xbox CPU is neither a P-III nor a Celeron, but something inbetween. It has only 128KB of L2 cache (like a Celeron, a real P-III has 256KB), but that cache is 8-way set associative (like a real P-III, a Celeron only has 4-way set associative cache).
The funniest part of it all is that these companies (JVC, in this case) are actually PAYING engineers and the like to implement these innovations.
JVC are not paying engineers to produce an innovation which allows the creation of uncopyable CDs. JVC are paying engineers to produce an innovation which can be sold to software companies for 20c to $1.00 per CD. In this endeavour, they are perfectly likely to succeed and prosper.
Yeah, I'm thinking it's a typo and they meant gigaflop.
Dude, a PS2 can do more than a gigaflop. (get the CPU and both VUs churning nicely and you'll almost hit 3 gigaflops..), so I don't think it was a typo. I'm pretty suspicious too of a claim that a 16-cored chip is going to do a teraflop (60+ gigaflops per core??) but I think it's their hyping, not their typing, that's at fault.
I think you'll find that what he meant is: a lot of people were claiming to have children (for the tax breaks you mention) when, in fact, they did not. When SS numbers started being demanded, they then had to stop claiming to have children. Hence there were "a lot less children" after this change. All the fake children went away.
Maybe it's changed in Windows XP or MacOS X. But for Windows 2000 and Redhat Linux 7.2 I have to install and run a separate program and laboriously pick out which files I want to burn and finally say "go".
When I had @Home, I had to agree to a one year contract if I wanted the installation fee to be waived. If I were still with them (which I'm not because they suck), I would remind them of the contract to provide unlimited access and that they can't raise the rate or implement limits until such contact was concluded.
I guarantee you that your contract clearly stated that they could raise prices or implement limits at any time whatsoever, with you having no say in it.
If you were lucky, it might say that in such a case, you can terminate the contract without penalty.
From the article: MacGeevy cited a recent UK raid on a DVD-R factory turning that was allegedly making copies of Spider Man and Star Wars: Episode II movies. The raid netted over 10,000 discs and 31 DVD burners.
See! We need these ID numbers! If we had them, we could find out who bought the original Spiderman and Episode II DVDs that were being copied in this DVD-R factory! We could trace the credit card records, find the person who purchased these DVDs, and more importantly, where he purchased them.
Come on, don't tell me you wouldn't love to find a shop selling these DVDs..?
I thought that too, but after looking around the Playstation2 section at CompUSA I noticed that there are a LOT of PC games (Half-life, No One Lives Forever, Baldurs Gate to name a few) that have been ported to the PS2. That makes me think that perhaps the hardware, or at least the PS2 API, is that different from a PC.
Baldurs Gate: Dark Alliance isn't a port, its a totally different game set in the same world. I don't know about No One Lives Forever, but as for Half Life, a game that old is probably simple enough to run well on a PS2 even if the port is crap.
There is no "PS2 API".. either you buy an engine from someone, or you write your own.
Bah! Just grabbed "(smr) Attack_of_the_clones (1of2).avi" (136,856k) from Grokster.. it's a fake, looks like "Showtime" with De Niro and Eddie Murphy. Which I kinda wanted to see, but y'know.
I do wonder about the mentality of people who keep these fakes shared out though. You want your bandwidth sucked up by people leeching them from you?
Slashdot Mirror
The doomsayers were right. USENET is a vast wasteland now. Ask your mommy and daddy what it used to be like before the Spammers destroyed it.
Maybe I'm just lucky with the 3 newsgroups I read, but none of them have more than maybe 1% spam. Literally: a couple of spams a week. They're not moderated newsgroups or anything, either - just one aus.* local group, one in the rec.games.* heirarchy, and one in comp.lang.*
The reports of Usenet's death are greatly exaggerated, if you ask me. I read news with Mozilla, so I don't even have a killfile! And yet, I'm perfectly happy.
Charged at $8 per gigabyte over the cap. Reliable estimates by people in the know estimate that Bell's actual costs for bandwidth are in the range of 50 cents to $1 per gigabyte.
Sucks to be you. Not. My Telstra ADSL account, touted as the model these American ISPs are talking about emulating, charges 13.9c per MEGAbyte over the cap - that works out to about $A142 per gig.
So you can understand that I'm very careful not to go over the limit. Of course, because I hate Telstra, come the last day of the month I'm downloading any old crap just to make sure I get as close to the limit as I can. Don't want to let them off providing less bandwidth than I'm paying for!
Seriously, if I could get a 19" LCD for $200-300 more than a 19" CRT with the same resolution, I'd do it.
Bear in mind that a 19" LCD will have a larger viewable area than a 19" CRT. Try comparing a 17" LCD to a 19" CRT, you'll probably find that they look much the same size.
I went from a 17" CRT to a 15" LCD (Hercules Prophetview), and it looks to be much the same size, and a hell of a lot easier on the eyes. It only does 1024x768 natively, but that's good enough for me, and running games at lower res actually looks excellent - the panel bilinearly filters it up to 1024x768 (no crappy pixel doubling), and I think 640x480 for instance looks better than it did on a CRT that could display it "properly".
It's not as bad as naming the golf game "kolf". Actually, wait a minute, no, they're both equally stupid names.
Hm, nice try, but here [vuurwerk.net] we can see that the N64 uses a MIPS R4300i, which is a 64-bit device.
I'm sorry, I neglected to mention that I actually live in a parallel universe where the current crop of Nintendo game consoles is made up of machines called "Gamecubes", and where "N64s" are actually the previous crop of Nintendo game consoles.
In my world, these "Gamecubes" use a CPU called Gekko, which is an IBM-developed PowerPC family chip, with 32-bit registers.
Aren't the current crop of Nintendo game consoles powered by a 64-bit MIPS?
No, they are powered by a 32-bit PowerPC-family chip.
Most variables have a minimum of 8 significant bits. The average length of a character string is in the 8-12 byte range (64-96 bits!)
..which almost always need to be operated on one character, i.e. 8 bits, at a time.
Third, what's this crap about it being "too expensive" to transfer 64-bits of data in from RAM? All modern processors have 64-bit wide data busses
..and said transfers are used to fill cache lines, not to read individual variables. With 64-bit variables, you fit half as many into each of those cache lines, and thus go crashing back to memory accesses twice as often.
The way I see it, Divx needs 3 things before it becomes a major threat to DVD.
1-Players capable of playing multiple soundtracks, for multiple languages and/or commentary.
Well, I don't know about 2 and 3, but you can do that one now - you just need to put the DivX video into an OGM (Ogg Multimedia) container rather than an AVI container. Then you can have multiple soundtracks (and they can be VBR audio too, whereas AVI only works properly with CBR).
That guy? I reckon this guy owes us an explanation!
800 MHz G4 is faster crunching the keys than a 2 GHz Athlon XP. I am reading that right?
Yes. I've never seen anything which shows off Altivec quite as well as RC5 cracking. There are hand optimised assembly cores for various CPUs in the d.net client, but the Altivec-enhanced G4 core pretty much destroys everything. I expect it's because Altivec has vastly more flexible shuffling instructions than MMX.
This by no means proves that a G4 is "better" than an Athlon, but it's interesting.
Four-digit punk. :)
(who's next!?)
Yep, you're retarded. :-) Nah, seriously, the pixel fill rate is 2600 (million pixels/sec) compared to 1100, 1200 and 880 for the three compared cards. So, yes, more than double that of the closest competition.
You're looking at the texel fillrate, which is also 2600 (million texels/sec) for the Radeon 9700, and is not blistering ahead of the competition (indeed, as you say, it's below one of them, the Matrox Parhelia).
But, as they say a few paragraphs down, it's better to get a high max texel fill rate from high clock speed and lots of pipes (as the Radeon 9700 does) rather than from lots of texture units per pipe (as the Parhelia does), because not every game is going to want to use 4 textures on every pixel. They all want to draw a helluvalot of pixels, though.
On a my P3-800 with 256mb of ram and a geforce 2 ti it ran like ass anytime there were more than 3 guys running around or if I was in a big room. Really pissed me off. And that was at 800x600, I had to turn it down from 1024x768 because it was unplayable.
News flash: if changing resolution improves performance, then your problem is that you're fillrate bound on the graphics card. Nothing to do with your CPU, nothing to do with "lazy/inefficient programming".
If you were getting the same crappy performance regardless of resolution, then you'd have a point.
"Why fix what ain't broken" is a damn good way to sum it up, IMO. This is coming from a guy who's perfectly happy running MacOS 8.6.1 on his G4, and WinME on his Windows boxes.
Many would say that you broke your Windows boxes when you "upgraded" to WinME from the far superior Win98.
While I applaud the open source community for producing such a high-quality competitor to MP3 as OGG, the real issue of getting people to switch still lies in hardware support and easy-to-use, CDDB compatible OGG CD-rip utilities.
Hardware support is the biggie. CD ripping is a non-issue - Exact Audio Copier and CDex, which are surely the two rippers most people are using, both support Ogg perfectly well (along with a swag of other non-MP3 formats).
I hate to question the logic of engineers, but it does seem pointless to me to keep the cache size the same, but make it 8-way set associative.
Well, I guess the important thing is that having the 128KB cache instead of 256KB saves money (obviously vital for a games console), whereas going with the 4-way cache instead of 8-way wouldn't save much if anything.
However, you do sometimes want to cache two address with the same lower bits, so to do this they came up with the idea of 2-way set associative, where you divide the memory in half, and make half of it a place to put the second-address with the same lower bits.
2-way set associative means that each piece of memory can be cached in two different places (and when it needs to be cached, some algorithm e.g. Least Recently Used will be used to decide which of those places it should be cached). This is unquestionably a good thing, it means you can avoid ping-ponging if two subroutines being called alternately both happen to map to the same cacheline. It certainly doesn't mean that you can only cache half as much stuff or anything bad like that.
8-way set associative is quite high.. probably high enough that as long as your code fits in cache, you won't get any evictions due to different chunks of code fighting over the same cachelines..
There can be a performance drawback to increasing n in an n-way set associative cache, as the increased complexity of the cache lookup logic means that it can be slower to retrieve a cacheline. I have no idea if there is a difference in practice between the P-III and Celeron L2 caches, though.
Pentium III ? More likely Celery^Hon according to this screenshot...
Actually the Xbox CPU is neither a P-III nor a Celeron, but something inbetween. It has only 128KB of L2 cache (like a Celeron, a real P-III has 256KB), but that cache is 8-way set associative (like a real P-III, a Celeron only has 4-way set associative cache).
Surely it's because Google's cache would allow people inside the Great Firewall to read all manner of banned web pages?
The funniest part of it all is that these companies (JVC, in this case) are actually PAYING engineers and the like to implement these innovations.
JVC are not paying engineers to produce an innovation which allows the creation of uncopyable CDs. JVC are paying engineers to produce an innovation which can be sold to software companies for 20c to $1.00 per CD. In this endeavour, they are perfectly likely to succeed and prosper.
Yeah, I'm thinking it's a typo and they meant gigaflop.
Dude, a PS2 can do more than a gigaflop. (get the CPU and both VUs churning nicely and you'll almost hit 3 gigaflops..), so I don't think it was a typo. I'm pretty suspicious too of a claim that a 16-cored chip is going to do a teraflop (60+ gigaflops per core??) but I think it's their hyping, not their typing, that's at fault.
I think you'll find that what he meant is: a lot of people were claiming to have children (for the tax breaks you mention) when, in fact, they did not. When SS numbers started being demanded, they then had to stop claiming to have children. Hence there were "a lot less children" after this change. All the fake children went away.
Maybe it's changed in Windows XP or MacOS X. But for Windows 2000 and Redhat Linux 7.2 I have to install and run a separate program and laboriously pick out which files I want to burn and finally say "go".
It has changed in MacOS X.
When I had @Home, I had to agree to a one year contract if I wanted the installation fee to be waived. If I were still with them (which I'm not because they suck), I would remind them of the contract to provide unlimited access and that they can't raise the rate or implement limits until such contact was concluded.
I guarantee you that your contract clearly stated that they could raise prices or implement limits at any time whatsoever, with you having no say in it.
If you were lucky, it might say that in such a case, you can terminate the contract without penalty.
From the article:
MacGeevy cited a recent UK raid on a DVD-R factory turning that was allegedly making copies of Spider Man and Star Wars: Episode II movies. The raid netted over 10,000 discs and 31 DVD burners.
See! We need these ID numbers! If we had them, we could find out who bought the original Spiderman and Episode II DVDs that were being copied in this DVD-R factory! We could trace the credit card records, find the person who purchased these DVDs, and more importantly, where he purchased them.
Come on, don't tell me you wouldn't love to find a shop selling these DVDs..?
I thought that too, but after looking around the Playstation2 section at CompUSA I noticed that there are a LOT of PC games (Half-life, No One Lives Forever, Baldurs Gate to name a few) that have been ported to the PS2. That makes me think that perhaps the hardware, or at least the PS2 API, is that different from a PC.
Baldurs Gate: Dark Alliance isn't a port, its a totally different game set in the same world. I don't know about No One Lives Forever, but as for Half Life, a game that old is probably simple enough to run well on a PS2 even if the port is crap.
There is no "PS2 API".. either you buy an engine from someone, or you write your own.
Bah! Just grabbed "(smr) Attack_of_the_clones (1of2).avi" (136,856k) from Grokster.. it's a fake, looks like "Showtime" with De Niro and Eddie Murphy. Which I kinda wanted to see, but y'know.
I do wonder about the mentality of people who keep these fakes shared out though. You want your bandwidth sucked up by people leeching them from you?