Being able to use p2p networks and downloading music are not the same thing.
Can people just be honest and stop saying "p2p networks are not about piracy." That's simply not true. The sole purpose for these networks to exist is to propagate pirated music or programs.
Anyways, the point is having enough bandwidth to download music and being allowed on P2P networks are not the same thing.
Personally I'd rather they just throttled bandwidth instead. That addresses the problem they have more directly.
One thing more disturbing is the crack down on people running servers [e.g. http]. I say if you have a connection you should be able todo anything non-commercial you want to with it.... that's another rant all together.
What businesses should do instead is embrace the new technology. I'm not the first to say it but you can't "uninvent" things.
You know life is finite but we adapt to that [mostly]. Why can't people adapt to digital copies?
I'll tell you what, I bet tons of P2P phreaks would love to have a high bandwidth source of quality mp3s, even if it were a monthly subscription thingy.
Being able to say "I want to hear this song now" instead of "connect, search, filter out crappy copies, filter out other songs renamed to what I want, etc..." is much better.
But leave it to stupid humans to hold onto the past and beat themselves in the head with it.
This same fallacy is the same that governs religious wars. Instead of embracing different cultures people say "new, kill it!" and there you have it.
I'd say we don't need nor want DRM. The businesses that buy into this hype are only setting themselves up for a fall, while scam artists [www.cloakware.com] are laughing it all the way to the bank.
Tom
Re:I hate to rehash an old argument.....
on
ISP Chief on Spam
·
· Score: 1
" bandwith and processor on BOTH ends"
The tokens he was probably refering to was HashCash. Only the clients have todo the work. The idea behind hashcash is that if a message doesn't have the token [sent as an extra header in the email] then the client can filter it out easily.
The eventual goal is to have 99% of all mass-email being ignored.
Tokens don't stop dedicated people but they certainly hurt mass-spamming. Tokens definately don't need much from the server. A mere header like
X-hashcash: 348957349573489573895738
is all you [the sending client] need to store in the email. As far as the SMTP server is concerned its just another header.
So in reality you could curb spam with the *SAME* SMTP servers already deployed.
When are people going to realize that DRM is just a sham and the only way to enforce it is through law. You can't make data that is observable uncopyiable. No matter how much you hype your magic [www.cloakware.com] fairy dust, bits that are observable are copyable.
We already have piracy laws so do we really need DRM laws?
I seriously doubt the "intel much faster" issue. First my 1.53Ghz AMD performs on mosts tests i've seen as good as a 1.8Ghz P4. Actually in some number theoretic stuff [e.g. with my libtomcrypt library] my P4 2.0Ghz [actually belongs to my parents] is no better than my Athlon.
The P4 only really can beat the Athlon in applications that are very specifically built for the P4. The rest of the available 99% of the software out there is not normally optimized well for it.
Also as to the depth, I don't doubt that a longer pipeline is a good idea, just the lack of multiple pipes means that you can only complete so many instructions per cycle. For instance, the fastest an intel P4 can complete
ADD EAX,23 ADD EBX,46 ADD ECX,69
is in 3 cycles. An athlon could [dunno if it does] complete that in just 1 cycle.
Obviously this is very contrived but serves the point well. The Athlon can retire multiple instructions per clock cycle.
The downside to the Athlon design is that certain "vector" or "long" instructions [using AMD speak] could take more than 1 cycle each whereas the P4 would still take one cycle each in most cases.
AMD cpus might be cheaper because of demand but I also think its just AMD being smarter and realizing that for "toys" spending 500$ on a CPU is not a good idea.
From what I have read alot of the transistor count of the P4 is the cache, an SRAM requires at least 5 transistors per bit, at 512KB that is 512*1024*8*5==20 million transistors right there. The Athlon has 384KB of cache which only requires 384*1024*8*5==15 million transistors. The higher count probably makes for hire failure rates as well as more problems making room for the design.
A lot of the success of AMDs' design is that they make *better* use of the transistors then Intel does. Having multiple pipelines [3 ALU, 2 LS and 3FPU from what I read] means that the AMD processor can execute quite a bit per cycle.
The Intel processor while having a deep pipeline can at most retire one instruction per cycle. The deep pipeline does allow the processor to break down bigger instructions to remove stalls as well as up the clock rate.
The AMD processor on the other hand can retire multiple instructions at the same time. It is in theory possible to finish 3 MMX/FPU/3dNOW instructions at the same time as 3 ALU instructions [but not realistic since there would be tons of dependencies].
The only big failing of the AMD design is a weak northbridge implementation by mobo companies like ASUS. For instance, people bitch that the Athlon is a hot sucker but really when it is idle it can get as cool as the P4. [My Athlon XP can go as low as +10C from room temperature]
But their cpu's already suck. Given the option I'd definately rather have an AMD processor than an Intel.
Most of Intel's problem is that they are going to long ends [literally!] to get a high Ghz design while not actually thinking in terms of MIPS. AMD went the other direction and made a processor which can execute quite a bit per cycle. The result is that AMD cpus are cheaper, have few transistors [roughly 7 million less] and can take the same power [*].
If Intel just scrapped the P4 "long pipeline" route and went to a "many multiple pipeline" route like AMD they'd rock.
Also the P4 doesn't have MMX+, 3DNOW [or the any additions to it].
The AMD XP has the full x86 set, MMX, MMX+, all of 3DNOW and the first SSE set of opcodes. This makes it a serious competitor.
[*] A tool called "vcool" can tell the mobo [asus mobos] to let the cpu goto sleep when IDLE. By default AMD processors only enter low power state when the processor is disconnected... Lowered my idle temp from 52C to 31C
A serious misconception is that more transistors lead to faster designs.
Compare the Athlon XP and P4 [37.5 and 44 Million transistors respectively ]. The XP is actually faster in virtually every test for a similarly clocked P4 despite having *fewer* transistors.
Perhaps that's true. I don't own a laptop but the last time I used one I noticed the temperature fairly soon after turning it on.
Either way its a stupid way to get hurt. Considering how rarely this injury occurs I'd be more suspect of his ability to reason then the faults of the human sensory system.
"Today on Let's Get a Clue!: We describe to MS lackeys that NVidia drivers are third party software. We then spend four hours describing what the above means."
Something I just realized. Did Linus write GCC? Did he write KDE? Did he write XMMS? did he write abiword? did he write infozip? did he write...
Linux distros are *completely* made up of third party software. So I find your sly about "are third party software" kinda moot.
Its upto the distro creator to choose quality combo of software [not every combo like RH].
Now I have to find a random string R where HASH(R) matches h for the lower K bits [you choose K] this is called a "collision".
Now I send the message M, to person T along with R. They verify it.
This can be used initially just to sort email so you can prioritize it [e.g. hash cashed email and non].
The trick is the bulk of the time is on the sender and it really hinders. If it takes them a second or two to find a K bit collision then they obviously can't spam millions of people a second.
My example of "nvidia drivers" is typical of OSS stuff you find for linux that was my point. Half-hazard "this works on my box" crap that never installs properly.
My point is that unless you go out of your way, all win32 apps use the same API which means a win2k app will work in win98 or winxp.
The same is *possible* for linux distro's but often not the case. One developer likes the Gnome WM, another KDE, another..., this developer likes QT version 2.3.1 not 2.3.2, etc...
I'm not making this shit up people. While I'm not the most investigative person on earth that's because I have other things todo (http://libtomcrypt.iahu.ca). When I download a an app I expect to build and install it in two commands "./configure ; make install ". Simple enough, but when it says "you need lib XYZ" then you get lib XYZ, but to build that you need lib ABC, etc... its a pain in the ass.
As for apache, clicking on that box "installs" it, doesn't set it up. Granted modding the conf file is easy I'm just saying that in windows its particularly easier.
Actually I installed RH 8.0 just 9 days ago. I reverted to Win2k after installing my nvidia drivers killed my X install [GLX wouldn't load]. I followed the nvidia instructions to the letter and it just didn't work.
Whats worse is I couldn't revert my X install to the old GLX driver.
As for Apache, it is an app a typical business student will use if they are running their own business. [I know I've worked with them].
Me sir, am not an idiot. I'm just not infinitely patient to put up with Linux roadblocks.
While I agree that purposely crashing another application is very bad behaviour it is *their* OS to do so with. Plus nobody is saying you can't release your own patch to undo the windows patch.
Plus the real audio client sucks ass. They should have just sold their codec to MSFT and be done with. That way we could have RA support in a decent client WMP.
True I haven't tried Mandrake but I've seen enough of Linux [e.g. RH] to know the "you don't have the millionth copy of this RPM to install this package circular logic" crap.
Personally I have no reason to leave Win2K since a) I didn't pay for my copy, b) works like a charm and c) have yet to have problems installing drivers or using any app.
What boggles my mind is why Distro should be 3 full CD's to begin with. If they just consolidated their fucking support libs/apps then it wouldn't be so big.
Last time I installed Linux [RH 8.0] I got three versions of QT, five different kernels, two copies of GCC, KDE/Gnome support libs, various versions of Motif, etc, etc...
Now I know every linux user likes having "their" version of a program but this is very useless for the "I just want to use the damn thing" user [e.g. me].
Choice be damned, when you have to install 1400 packages [4GB] to get your RH install to work properly without any depend bitching there are serious problems.
If Mandrake is completely the opposite then maybe I should go check it out some time but so far I'm not that inclined.
"1. M$ is scared shitless of Linux. They have no real strategy to deal with something that even they know is more stable and secure, and know they can't compete on price."
While MS costs more to "buy", linux I'd say costs more to install. Almost anyone can setup and use a MS windows platform. Compariatively noone can install and use a linux distro. Ask some business student to install Apache [when they aren't that computer literal to begin with] is fun:-)
The rest of your post is fairly typical of MS-hate speak.
We have law makers introducing anti-terorism laws [e.g. patriot act et al.] You have law makers introducing new adjustments to miranda, you have law makers trying to break up a computer firm in a tough economic time.
Are any of these politicians actually working for the good of the public they represent?
Suppose they manage to shut down MS [or severely disrupt it]. What comes of that? 1000s of people lose their jobs.
I've been saying this for a few years now. The best way to "adjust" MSFT "way of business" is to hurt them fairly. Sell a better product. Sure not everyone will jump off the MS boat right away but being impatient and shutting them down via anti-trust is hardly the way to prove a point.
I mean at this stage for the average desktop user the average linux distro is hardly useful. Even RH 8.0 has some severe problems [like installing nvidia drivers can kill your X install just by following the instructions!].
By making MSFT illegal and leaving linux as the only option you'd actually be hurting the industry, not helping it.
When linux distros actually compete with Windows [e.g. in a meaningful sense, having 1500 packages on 3 CD's is not "competition" when installing a GFX driver can kill the install] then we'll see the beginning of the demise of Windows.
Slashdot Mirror
Being able to use p2p networks and downloading music are not the same thing.
Can people just be honest and stop saying "p2p networks are not about piracy." That's simply not true. The sole purpose for these networks to exist is to propagate pirated music or programs.
Anyways, the point is having enough bandwidth to download music and being allowed on P2P networks are not the same thing.
Personally I'd rather they just throttled bandwidth instead. That addresses the problem they have more directly.
One thing more disturbing is the crack down on people running servers [e.g. http]. I say if you have a connection you should be able todo anything non-commercial you want to with it.... that's another rant all together.
Tom
er, guns are supposed to be locked separately from the ammunition anyways.
.223 rifle] but I have no love for the "I store me gun under be bed" types either.
So your point is moot.
Either you dangerously store your weapons or they're of no use to you anyways [in that situation].
Don't get me wrong, I love shooting too [.40 pistol and
Tom
True, but I'd rather not let people have guns in their homes to begin with. Sure own a gun, *at a club!*.
Tom
What businesses should do instead is embrace the new technology. I'm not the first to say it but you can't "uninvent" things.
You know life is finite but we adapt to that [mostly]. Why can't people adapt to digital copies?
I'll tell you what, I bet tons of P2P phreaks would love to have a high bandwidth source of quality mp3s, even if it were a monthly subscription thingy.
Being able to say "I want to hear this song now" instead of "connect, search, filter out crappy copies, filter out other songs renamed to what I want, etc..." is much better.
But leave it to stupid humans to hold onto the past and beat themselves in the head with it.
This same fallacy is the same that governs religious wars. Instead of embracing different cultures people say "new, kill it!" and there you have it.
I'd say we don't need nor want DRM. The businesses that buy into this hype are only setting themselves up for a fall, while scam artists [www.cloakware.com] are laughing it all the way to the bank.
Tom
" bandwith and processor on BOTH ends"
The tokens he was probably refering to was HashCash. Only the clients have todo the work. The idea behind hashcash is that if a message doesn't have the token [sent as an extra header in the email] then the client can filter it out easily.
The eventual goal is to have 99% of all mass-email being ignored.
Tokens don't stop dedicated people but they certainly hurt mass-spamming. Tokens definately don't need much from the server. A mere header like
X-hashcash: 348957349573489573895738
is all you [the sending client] need to store in the email. As far as the SMTP server is concerned its just another header.
So in reality you could curb spam with the *SAME* SMTP servers already deployed.
Tom
When are people going to realize that DRM is just a sham and the only way to enforce it is through law. You can't make data that is observable uncopyiable. No matter how much you hype your magic [www.cloakware.com] fairy dust, bits that are observable are copyable.
We already have piracy laws so do we really need DRM laws?
Ok there is the "right" and "reality" of the situation.
In essence, yeah, broadband users were scammed into thinking they have a high speed connection 24/7.
That's not the reality of the situation.
You have three options though.
1. Sue your provider.
2. Use all the bw you want and get charged more in return.
3. Be reasonable with your bw usage as if you're sharing it with others [hint hint].
We all know that #1 is the "right" action but #3 is more realistic and in the long run more likely to result in a "good" outcome than #1.
Tom
What redmond scum?
Besides I figure this is a way for ISPs to save bandwidth cost... if all the users are rooted and can't go on P2P networks and such all the better.
Tom
More specifically the N64 is a ~95Mhz 64-bit RISC core [dunno which series] with 4MB of ram.
:-)
When is the last time any gaming PC had 4MB of ram?
I seriously doubt the "intel much faster" issue. First my 1.53Ghz AMD performs on mosts tests i've seen as good as a 1.8Ghz P4. Actually in some number theoretic stuff [e.g. with my libtomcrypt library] my P4 2.0Ghz [actually belongs to my parents] is no better than my Athlon.
The P4 only really can beat the Athlon in applications that are very specifically built for the P4. The rest of the available 99% of the software out there is not normally optimized well for it.
Also as to the depth, I don't doubt that a longer pipeline is a good idea, just the lack of multiple pipes means that you can only complete so many instructions per cycle. For instance, the fastest an intel P4 can complete
ADD EAX,23
ADD EBX,46
ADD ECX,69
is in 3 cycles. An athlon could [dunno if it does] complete that in just 1 cycle.
Obviously this is very contrived but serves the point well. The Athlon can retire multiple instructions per clock cycle.
The downside to the Athlon design is that certain "vector" or "long" instructions [using AMD speak] could take more than 1 cycle each whereas the P4 would still take one cycle each in most cases.
AMD cpus might be cheaper because of demand but I also think its just AMD being smarter and realizing that for "toys" spending 500$ on a CPU is not a good idea.
Tom
From what I have read alot of the transistor count of the P4 is the cache, an SRAM requires at least 5 transistors per bit, at 512KB that is 512*1024*8*5==20 million transistors right there. The Athlon has 384KB of cache which only requires 384*1024*8*5==15 million transistors. The higher count probably makes for hire failure rates as well as more problems making room for the design.
A lot of the success of AMDs' design is that they make *better* use of the transistors then Intel does. Having multiple pipelines [3 ALU, 2 LS and 3FPU from what I read] means that the AMD processor can execute quite a bit per cycle.
The Intel processor while having a deep pipeline can at most retire one instruction per cycle. The deep pipeline does allow the processor to break down bigger instructions to remove stalls as well as up the clock rate.
The AMD processor on the other hand can retire multiple instructions at the same time. It is in theory possible to finish 3 MMX/FPU/3dNOW instructions at the same time as 3 ALU instructions [but not realistic since there would be tons of dependencies].
The only big failing of the AMD design is a weak northbridge implementation by mobo companies like ASUS. For instance, people bitch that the Athlon is a hot sucker but really when it is idle it can get as cool as the P4. [My Athlon XP can go as low as +10C from room temperature]
Tom
But their cpu's already suck. Given the option I'd definately rather have an AMD processor than an Intel.
Most of Intel's problem is that they are going to long ends [literally!] to get a high Ghz design while not actually thinking in terms of MIPS. AMD went the other direction and made a processor which can execute quite a bit per cycle. The result is that AMD cpus are cheaper, have few transistors [roughly 7 million less] and can take the same power [*].
If Intel just scrapped the P4 "long pipeline" route and went to a "many multiple pipeline" route like AMD they'd rock.
Also the P4 doesn't have MMX+, 3DNOW [or the any additions to it].
The AMD XP has the full x86 set, MMX, MMX+, all of 3DNOW and the first SSE set of opcodes. This makes it a serious competitor.
[*] A tool called "vcool" can tell the mobo [asus mobos] to let the cpu goto sleep when IDLE. By default AMD processors only enter low power state when the processor is disconnected... Lowered my idle temp from 52C to 31C
Tom
A serious misconception is that more transistors lead to faster designs.
Compare the Athlon XP and P4 [37.5 and 44 Million transistors respectively ]. The XP is actually faster in virtually every test for a similarly clocked P4 despite having *fewer* transistors.
Tom
Basically what Adobe is saying.
- We want elcomsoft to not release software that breaks our unbreakable security software.
Funny.
You'd think Adobe would either just step up and admit they ain't got shit, or fix their software. Mostly realize you can't make bits uncopyable...
Tom
Perhaps that's true. I don't own a laptop but the last time I used one I noticed the temperature fairly soon after turning it on.
Either way its a stupid way to get hurt. Considering how rarely this injury occurs I'd be more suspect of his ability to reason then the faults of the human sensory system.
Tom
A lot of things were too late for that guy. Like common sense.
Next you know he'll sue frying pan makers because he likes eating bacon right off the pan which he rests on his lap...
If its hot, don't touch it... simple rule.
Tom
"Today on Let's Get a Clue!: We describe to MS lackeys that NVidia drivers are third party software. We then spend four hours describing what the above means."
...
Something I just realized. Did Linus write GCC? Did he write KDE? Did he write XMMS? did he write abiword? did he write infozip? did he write
Linux distros are *completely* made up of third party software. So I find your sly about "are third party software" kinda moot.
Its upto the distro creator to choose quality combo of software [not every combo like RH].
Tom
Hash cash is the answer.
Ok I want to send letter M to person T, I find
h = HASH(M || T)
Now I have to find a random string R where HASH(R) matches h for the lower K bits [you choose K] this is called a "collision".
Now I send the message M, to person T along with R. They verify it.
This can be used initially just to sort email so you can prioritize it [e.g. hash cashed email and non].
The trick is the bulk of the time is on the sender and it really hinders. If it takes them a second or two to find a K bit collision then they obviously can't spam millions of people a second.
Tom
My example of "nvidia drivers" is typical of OSS stuff you find for linux that was my point. Half-hazard "this works on my box" crap that never installs properly.
..., this developer likes QT version 2.3.1 not 2.3.2, etc...
My point is that unless you go out of your way, all win32 apps use the same API which means a win2k app will work in win98 or winxp.
The same is *possible* for linux distro's but often not the case. One developer likes the Gnome WM, another KDE, another
I'm not making this shit up people. While I'm not the most investigative person on earth that's because I have other things todo (http://libtomcrypt.iahu.ca). When I download a an app I expect to build and install it in two commands "./configure ; make install ". Simple enough, but when it says "you need lib XYZ" then you get lib XYZ, but to build that you need lib ABC, etc... its a pain in the ass.
As for apache, clicking on that box "installs" it, doesn't set it up. Granted modding the conf file is easy I'm just saying that in windows its particularly easier.
Tom
Actually I installed RH 8.0 just 9 days ago. I reverted to Win2k after installing my nvidia drivers killed my X install [GLX wouldn't load]. I followed the nvidia instructions to the letter and it just didn't work.
Whats worse is I couldn't revert my X install to the old GLX driver.
As for Apache, it is an app a typical business student will use if they are running their own business. [I know I've worked with them].
Me sir, am not an idiot. I'm just not infinitely patient to put up with Linux roadblocks.
Tom
While I agree that purposely crashing another application is very bad behaviour it is *their* OS to do so with. Plus nobody is saying you can't release your own patch to undo the windows patch.
Plus the real audio client sucks ass. They should have just sold their codec to MSFT and be done with. That way we could have RA support in a decent client WMP.
Tom
True I haven't tried Mandrake but I've seen enough of Linux [e.g. RH] to know the "you don't have the millionth copy of this RPM to install this package circular logic" crap.
Personally I have no reason to leave Win2K since a) I didn't pay for my copy, b) works like a charm and c) have yet to have problems installing drivers or using any app.
What boggles my mind is why Distro should be 3 full CD's to begin with. If they just consolidated their fucking support libs/apps then it wouldn't be so big.
Last time I installed Linux [RH 8.0] I got three versions of QT, five different kernels, two copies of GCC, KDE/Gnome support libs, various versions of Motif, etc, etc...
Now I know every linux user likes having "their" version of a program but this is very useless for the "I just want to use the damn thing" user [e.g. me].
Choice be damned, when you have to install 1400 packages [4GB] to get your RH install to work properly without any depend bitching there are serious problems.
If Mandrake is completely the opposite then maybe I should go check it out some time but so far I'm not that inclined.
Tom
"1. M$ is scared shitless of Linux. They have no real strategy to deal with something that even they know is more stable and secure, and know they can't compete on price."
:-)
While MS costs more to "buy", linux I'd say costs more to install. Almost anyone can setup and use a MS windows platform. Compariatively noone can install and use a linux distro. Ask some business student to install Apache [when they aren't that computer literal to begin with] is fun
The rest of your post is fairly typical of MS-hate speak.
Tom
We have law makers introducing anti-terorism laws [e.g. patriot act et al.] You have law makers introducing new adjustments to miranda, you have law makers trying to break up a computer firm in a tough economic time.
Are any of these politicians actually working for the good of the public they represent?
Suppose they manage to shut down MS [or severely disrupt it]. What comes of that? 1000s of people lose their jobs.
I've been saying this for a few years now. The best way to "adjust" MSFT "way of business" is to hurt them fairly. Sell a better product. Sure not everyone will jump off the MS boat right away but being impatient and shutting them down via anti-trust is hardly the way to prove a point.
I mean at this stage for the average desktop user the average linux distro is hardly useful. Even RH 8.0 has some severe problems [like installing nvidia drivers can kill your X install just by following the instructions!].
By making MSFT illegal and leaving linux as the only option you'd actually be hurting the industry, not helping it.
When linux distros actually compete with Windows [e.g. in a meaningful sense, having 1500 packages on 3 CD's is not "competition" when installing a GFX driver can kill the install] then we'll see the beginning of the demise of Windows.
Tom
Wrong. With ASUS you can flash the mobo BIOS while windows is running.
Tom