Small niggle with the article
on
2.2 vs 2.4
·
· Score: 3
A reasonably well written article with one small problem:
it's pretty rare for a IDE drive to completely saturate all 33 MB of the IDE pipe
Whilst it is true that no drive will supply a sustained 33mb/sec, this forgets the overhead inherent in the EIDE protocol itself, which can be up to 30% of the bandwidth down the EIDE bus. Also remember this is per drive too.
Even if Sega *were* to halt Dreamcast production - which they've denied, the Dreamcast is getting a position as *the* console for hackers, so isn't about to die completely just yet. I just got my serial cable and have been playing about, and there's a lot of scope for doing some really kool stuff. The fact that GCC can compile SH-4 well is a big help. If you want to get into it, check out sites like boob.
Anyway, remember that it's all just a rumour at the moment, although we all know Sega's been having serious problems.
It would seem rather odd for this to come on a day when Sega's share price has increased by 200 Yen a share though.. dirty tactics perhaps?
Actually, here in the Uk at least most providers charge per second rather than per K. Some providers, mine included are now starting to offer 'free wap', which suddenly makes the service much more attractive. I pay £10 per month for a contract which includes unlimited WAP usage.
I have to say I was wondering why any company would buy licenses for this
amount. I work for a realtivly large multi-national OEM, and its something
like £5 per WinMe license (they're also allowed to press their own CDs
and print their own manual copies which makes things cheaper, that's not
included in this price though).
It seems to me this may be a last-ditch
publicity stunt by a company set to go under..
2.2.18 and 2.4.0 both have excellent USB support. The issue is in the drivers for USB devices. Complaining about 'rudimentary' USB support is like complaining about 'rudimentary' PCI support because your Winmodem doesn't work.
But, I suppose, if you're totally willing to trust your government with everything, ie. a cradle to grave welfare state, you might was well trust the nanny state to protect your privacy.
Here here, god forbid anyone get the idea that government could be part of society rather than an evil entity trying to control it. Of course, I can see where your confusion comes from - our government in its current state at times is an evil entity trying to control it. So how about this - why don't we bring government back to the people. Why not give small, locally elected bodies a great deal more power? I think that's something most of those who espouse the views contained within your message (read: the far right) would agree with. That way, a cradle to grave welfare state could be the local community helping itself, and not having to do stupid things like checking that the insurance you bought (you know, from that faceless corporation that is trying to control you and is collecting information on you) covers this particular disease or treatment method.
The other side of the coin is to take responsibility for your own protection by arming yourself, against intruders or a dangerous government.
Take arms against the government? Rather like the Bolshevik revolution then?
Oh, and though you'd hate to admit it, those small-but powerful local governments are called Soviets, a Marxist-Lenninist idea. Its a shame Stalin was more interested in state-run capitalism than Communism, and so didn't implement them.
<sigh> When will people realise that the government really needn't be their enemy? If only we are prepared to stand up and change it for the better, we can all be the government. Advocating random shootings really helps no-one except gun-manufacturers.
Except that our computers, as directed by the overlords at Microsoft, are installed with the OPK routine, which makes the customer enter their own product ID and select which country they're in.
Mule?
Does anyone else object to the way the article speaks about Moores law in an overly matter-of-fact manner, implying that it is just as much a law of Physics as Boyle's Law? A more careful journalist, rather than writing:
Moores Law, formulated by Intel co-founder Gordon Moore, states that the number of transistors a chip can hold will double every 18 to 24 months, as transistor size shrinks.
A more careful journalist would hopefully have written:
Moores Law, formulated by Intel co-founder Gordon Moore,
speculates that the number of transistors a chip can hold will double every 18 to 24 months, as transistor size shrinks.
Being a poor student I supplement my income by doing technical support for an OEM. Unfortunately, this OEM sells some of its computers through QVC. Everyone who works there dreads the 'QVC Customer' announcement before a call comes through, because we all know it'll be some clueless mother-fscker who will say things like 'Double-click? What does that mean?', 'Which one's the right mouse button', 'Where do I click the escape key' and 'I plugged all the cords into the telivision but it won't come on!'. In fact the best you can hope for from a QVC customer is the old 'My keyboard is broken - the double quotes above the 2 comes up as an @ sign and the pound sign doesn't work either!' (incase you hadn't guessed I'm in the UK).
Argh!
On the other side this goes wrong: if I want to make a graphical application, what toolkit should I use. I can choose between gtk, qt, gnome enhanced gtk, motif/lesstif or plain Xt or Xp. Tell me?
I run Gnome. I can run compile and run applications written using qt, qt2, gtk, gtk++, motif, lesstif, Xt, Xp, and any other library. Because everything is open, you can essentially have your cake and eat it.
When it comes down to it, the sort of people that currently run MS operating systems wouldn't know what a programming lanugage was, let alone a toolkit. As long as they click the pretty pictures and it does what they expect (ie what it did last time, or what the bored-of-their-skull tech support guy told them it would do) then they really don't care that it might look slightly different.
That's not to say they don't like shiny/pretty things of course; but I think we can safely say with KDE2 and Ximian we've got prettier desktops than them:)
The point of all the framebuffer development going on at the moment is largely for embedded device support, where Linux has extremely good market share already.
Is that (from my reading of the article at least, I may be wrong) they require HDTVs. HDTV isn't out in Europe, and isn't likely to be any time soon (we'll just stick to our lovely 100hz PAL Widescreen Anamorphic Digital TVs thanks:)).
Despite what many in North America may believe, the US doesn't rule the world.
You are speaking from a UNIX perspective. I, however, speak from a free software perspective. Apple has done nothing to help the free software community, in fact with its licensing of the 1-Click Patent it has done much to harm freedom. Apple has used no Linux technology in its new OS, which is BSD based. Most importantly none of it, even Darwin is free..
Face it, does anyone think that Apple would behave much different if they had the monopoly (bear in mind their treatment of clone makers)? Nope, neither do I.
IBM and Intel say that The Register's story mistakenly assumes that CPRM is intended for fixed hard disks, whereas it's only intended for removable media. Is this true?
Not if you examine the ATA extensions under consideration closely.
FACT: The CPRM ATA call interface requires information that standard ATA hard disks need, but that packet based removable ATAPI drives such as Zip and Jaz drives,don't: such as sector start and offset information. If the CPRM proposal under consideration by T.13 was for packet-based ATAPI drives, it wouldn't need this information.
FACT: We know of only one removable ATA drive: Castlewood's Orb. All others use ATAPI, or media-specific extensions on top of ATA (as with IBM's Microdrives) - that don't require extensions to the ATA command set.
From our conversations with the people behind the proposal, and public documents released by the T.13 committee, we'll agree that their focus to date has been on removable drives, and it's apparent that not all of the consequences of CPRM in fixed-drives have been discussed.
But unforeseen or not - and despite public protestations of their good intentions - the 4C Entity is delivering a solution tailor-made for fixed disk ATA drives, and building right into the specification for industry standard fixed drives. This is indisputable.
- clear 1X1 pixel gifs used for spacing with alt tags that say "spacer" - doing typesetting with 1X1 pixel transparent gifs is a kludge that adds a lot of excess html to your docs
Actually, it's sometimes a necessary evil to get everything looking 'right'. Creating a gap that small isn't possible on all browsers in any other way.
I'll agree with your other points though..
AFAIK (I used to do 'Research and Development' for a large OEM, but this was a year or so ago now, so things may have changed back for the good. I doubt it tho..)
The Celeron is fully capable of running SMP, however since ~500mhz (can't remember exactly when) Intel have stopped them from working together. There may be a hardware workaround (I believe there was some talk of doing things with slockets when slot-1 was still about), but that's the official position.
Copy protection hard drive plan nixes free software - RMS
By: Andrew Orlowski in San Francisco
Posted: 23/12/2000 at 01:07 GMT
Richard Stallman says that plans to put content control into industry standard hardware pose a threat to the adoption of free software.
Proposals have been made to add CPRM (Content Protection for Removable Media) into the ATA hard disk specification, we reported on Wednesday. CPRM originates from the the 4C Entity and licensing is administered by License Management International, LLC, which also administers the CSS license.
"This resembles CSS and e-Books: it is another plan to impose additional power over people who use published information, on behalf of those who hope to control the power," he writes in emails to The Register.
"This plan seems to pose a threat to free operating systems. We will surely not be authorized in the US to implement free software to access any of the centrally-controlled data. So a free GNU/Linux system won't be able to do it."
"If users accept the domination of centrally-controlled data, free software faces two dangers, each worse than the other: that users will reject GNU/Linux because it doesn't support the central control over access to these data, or that they will reject free versions of GNU/Linux for versions "enhanced" with proprietary software that support it. Either outcome will be a grave loss for our freedom."
"We must hope that some countries refuse to pass laws to prohibit free software such as DeCSS, so that some part of the world can publish the software that will keep freedom alive, underground, in the rest of the world."
Stallman also highlights the term "copy protection". "The word 'protection'... tries to disguise obstructionism and rampant power as an attempt to keep a program or book or song safe from harm. It is a propaganda word."
Indeed: it's a euphemism as incongruous as down-sizing or friendly fire. As an alternative, we quite like "copy control". But if you have snappier suggestions, we'd like to hear them.
Nice to hear about the smaller players, especially when they're making inexpensive integrated hardware my mom would like to use.
VIA a small player??? They make chipsets for a good 50% of motherboards (all the AMD ones and quite a few Pentium ones too), they make their own motherboards under the name of FIC, and they make memory too. VIA are anything but small, they're bigger than AMD in terms of capitol if not in terms of profile...
As a British subject (remember, we are not citizens), I say we all write to our respective and MPs and request that British Telecom be disallowed to use the word British in its name or any of its literature, for we own this word!
Firstly, I do disagree with you, and I also have mod points, but it's far more constructive to say why I disagree with you rather than simply moderate your post down. Hopefully other moderators will take this position too.
Anyway, let me address your points in turn
Has anyone considered the cost of encrypting the majority of Internet traffic? I always encrypt terminal sessions, but the cost of servers encrypting web traffic would be very high. I wouldn't mind; my PC has enough power to encrypt all MY traffic, but how about a busy web server? How much new hardware would slashdot need to support encrypting everything? If slashdot is going to blow that stupid smoke, then slashdot should encrypt all its traffic.
Incase you hadn't noticed hardware companies are producing faster and faster processors all the time, and the price of this increased technology is falling. We are very rapidly approaching (in fact it could be argued were are there) a time in which CPU speed routinely exceeds requirements. Encrypting data produces no overhead in terms of bandwidth, only in terms of CPU time. I'd wager that if the slashdot cru were to go out and replace the hardware they currently have, spending exactly the same amount as they did last time they upgraded, the extra CPU power they'd gain would be enough to add SSL.
Second, is his explicit assumption that Linux is the best thing available, and that free software is always better than proprietary software in quality. Linux does have a lot of good points, and in some cases is the best solution. MS Windows has strengths also. Sometimes Microsoft solutions solve a problem better. Sometimes one of the other systems he ignores is better.
What you say at first is incorrect; the interviewer says Linux is the best thing avaliable, he actually says that free software is the best thing avaliable, which happens to include Linux. The point that is being made is that even if the Microsoft implementation of something were to be technically better (and I'm not going to get into whether or not that is the case), the fact that it is not free makes it inherently worse. Microsoft may make something better, but time does not stand still. If something is free, anyone is free to improve it, and with the potential billions avaliable to improve it, there is no way proprietary software can compete.
Mandrake comes with a seriously non standard, patched kernel. (They are actually more than likely using fairly recent emu10k1 drivers, which were probably from opensource.creative.com), they also more than likely patched for usb support and other such little things.
Yup, in fact that should be kernels, it actually comes with two - a 'standard version' which is pretty heavily patched and a 'secure' version with improvements like a non-executable stack and ps showing only the current user's processes. Last time I checked both were still at 2.2.16 (although I'm still using 7.1, don't know if there are more updates for 7.2).
Yea, but there's no difference between using a Live! and a cheap soundcard with clean-sounding circitry. Linux doesn't use any of the acceleration features on the sound card, and for the money, there are a lot of cards that have cleaner sound, sans 3D hardware.
I've yet to see a cheap soundcard with anything like as low a S/N ratio as the SB Live! or a DAC even approaching the same quality. I doubt such a beast exists. Of course, unless you've got it hooked up to a decent amp/speakers then you won't hear the differenece.
Is that compiler optimisation is probably going to count for more and more as Intel wring every last bit of performance out of x86. Linux distros like Mandrake, which currently comes in a version optimised for P5, could potentially have big performance benefits if you get the version compiled for your specific processor. Of course, the really good news is that Intel et al will have to take quite an active role in GCC development if they want to make their processor look great under free operating systems.
Unfortunately, what it could potentially mean though is that if Intel were to do some sort of special deal with a proprietry OS maker (MS for example) they could make that OS run far faster than any others, simply because it'd be compiled with a better optimised compiler.
--
Piracy is a vicitmless crime, like punching someone in the dark.
Slashdot Mirror
Whilst it is true that no drive will supply a sustained 33mb/sec, this forgets the overhead inherent in the EIDE protocol itself, which can be up to 30% of the bandwidth down the EIDE bus. Also remember this is per drive too.
--
Even if Sega *were* to halt Dreamcast production - which they've denied, the Dreamcast is getting a position as *the* console for hackers, so isn't about to die completely just yet. I just got my serial cable and have been playing about, and there's a lot of scope for doing some really kool stuff. The fact that GCC can compile SH-4 well is a big help. If you want to get into it, check out sites like boob.
Anyway, remember that it's all just a rumour at the moment, although we all know Sega's been having serious problems.
It would seem rather odd for this to come on a day when Sega's share price has increased by 200 Yen a share though.. dirty tactics perhaps?
--
Actually, here in the Uk at least most providers charge per second rather than per K. Some providers, mine included are now starting to offer 'free wap', which suddenly makes the service much more attractive. I pay £10 per month for a contract which includes unlimited WAP usage.
--
I have to say I was wondering why any company would buy licenses for this amount. I work for a realtivly large multi-national OEM, and its something like £5 per WinMe license (they're also allowed to press their own CDs and print their own manual copies which makes things cheaper, that's not included in this price though).
It seems to me this may be a last-ditch publicity stunt by a company set to go under..
--
2.2.18 and 2.4.0 both have excellent USB support. The issue is in the drivers for USB devices. Complaining about 'rudimentary' USB support is like complaining about 'rudimentary' PCI support because your Winmodem doesn't work.
--
Here here, god forbid anyone get the idea that government could be part of society rather than an evil entity trying to control it.
Of course, I can see where your confusion comes from - our government in its current state at times is an evil entity trying to control it. So how about this - why don't we bring government back to the people. Why not give small, locally elected bodies a great deal more power? I think that's something most of those who espouse the views contained within your message (read: the far right) would agree with. That way, a cradle to grave welfare state could be the local community helping itself, and not having to do stupid things like checking that the insurance you bought (you know, from that faceless corporation that is trying to control you and is collecting information on you) covers this particular disease or treatment method.
Take arms against the government? Rather like the Bolshevik revolution then?
Oh, and though you'd hate to admit it, those small-but powerful local governments are called Soviets, a Marxist-Lenninist idea. Its a shame Stalin was more interested in state-run capitalism than Communism, and so didn't implement them.
<sigh>
When will people realise that the government really needn't be their enemy? If only we are prepared to stand up and change it for the better, we can all be the government. Advocating random shootings really helps no-one except gun-manufacturers.
--
Except that our computers, as directed by the overlords at Microsoft, are installed with the OPK routine, which makes the customer enter their own product ID and select which country they're in.
Mule?
--
A more careful journalist would hopefully have written:
--
Being a poor student I supplement my income by doing technical support for an OEM. Unfortunately, this OEM sells some of its computers through QVC.
Everyone who works there dreads the 'QVC Customer' announcement before a call comes through, because we all know it'll be some clueless mother-fscker who will say things like 'Double-click? What does that mean?', 'Which one's the right mouse button', 'Where do I click the escape key' and 'I plugged all the cords into the telivision but it won't come on!'.
In fact the best you can hope for from a QVC customer is the old 'My keyboard is broken - the double quotes above the 2 comes up as an @ sign and the pound sign doesn't work either!' (incase you hadn't guessed I'm in the UK).
Argh!
</rant>
--
On the other side this goes wrong: if I want to make a graphical application, what toolkit should I use. I can choose between gtk, qt, gnome enhanced gtk, motif/lesstif or plain Xt or Xp. Tell me?
:)
I run Gnome. I can run compile and run applications written using qt, qt2, gtk, gtk++, motif, lesstif, Xt, Xp, and any other library. Because everything is open, you can essentially have your cake and eat it.
When it comes down to it, the sort of people that currently run MS operating systems wouldn't know what a programming lanugage was, let alone a toolkit. As long as they click the pretty pictures and it does what they expect (ie what it did last time, or what the bored-of-their-skull tech support guy told them it would do) then they really don't care that it might look slightly different.
That's not to say they don't like shiny/pretty things of course; but I think we can safely say with KDE2 and Ximian we've got prettier desktops than them
The point of all the framebuffer development going on at the moment is largely for embedded device support, where Linux has extremely good market share already.
--
Is that (from my reading of the article at least, I may be wrong) they require HDTVs. HDTV isn't out in Europe, and isn't likely to be any time soon (we'll just stick to our lovely 100hz PAL Widescreen Anamorphic Digital TVs thanks :)).
Despite what many in North America may believe, the US doesn't rule the world.
--
You are speaking from a UNIX perspective. I, however, speak from a free software perspective. Apple has done nothing to help the free software community, in fact with its licensing of the 1-Click Patent it has done much to harm freedom. Apple has used no Linux technology in its new OS, which is BSD based. Most importantly none of it, even Darwin is free..
Face it, does anyone think that Apple would behave much different if they had the monopoly (bear in mind their treatment of clone makers)? Nope, neither do I.
--
From The Register:
IBM and Intel say that The Register's story mistakenly assumes that CPRM is intended for fixed hard disks, whereas it's only intended for removable media. Is this true? Not if you examine the ATA extensions under consideration closely. FACT: The CPRM ATA call interface requires information that standard ATA hard disks need, but that packet based removable ATAPI drives such as Zip and Jaz drives,don't: such as sector start and offset information. If the CPRM proposal under consideration by T.13 was for packet-based ATAPI drives, it wouldn't need this information. FACT: We know of only one removable ATA drive: Castlewood's Orb. All others use ATAPI, or media-specific extensions on top of ATA (as with IBM's Microdrives) - that don't require extensions to the ATA command set. From our conversations with the people behind the proposal, and public documents released by the T.13 committee, we'll agree that their focus to date has been on removable drives, and it's apparent that not all of the consequences of CPRM in fixed-drives have been discussed. But unforeseen or not - and despite public protestations of their good intentions - the 4C Entity is delivering a solution tailor-made for fixed disk ATA drives, and building right into the specification for industry standard fixed drives.
This is indisputable.
Now ask yourself, why is it there?
--
- clear 1X1 pixel gifs used for spacing with alt tags that say "spacer" - doing typesetting with 1X1 pixel transparent gifs is a kludge that adds a lot of excess html to your docs
Actually, it's sometimes a necessary evil to get everything looking 'right'. Creating a gap that small isn't possible on all browsers in any other way.
I'll agree with your other points though..
--
AFAIK (I used to do 'Research and Development' for a large OEM, but this was a year or so ago now, so things may have changed back for the good. I doubt it tho..) The Celeron is fully capable of running SMP, however since ~500mhz (can't remember exactly when) Intel have stopped them from working together. There may be a hardware workaround (I believe there was some talk of doing things with slockets when slot-1 was still about), but that's the official position.
--
Recording and publishing companies don't vote . The people who are affected by the schemes these companies are pushing do vote .
Unfortunately, whilst they don't vote, they do spend millions of dollars lobbying congress.
--
Copy protection hard drive plan nixes free software - RMS By: Andrew Orlowski in San Francisco Posted: 23/12/2000 at 01:07 GMT
Richard Stallman says that plans to put content control into industry standard hardware pose a threat to the adoption of free software.
... tries to disguise obstructionism and rampant power as an attempt to keep a program or book or song safe from harm. It is a propaganda word."
Proposals have been made to add CPRM (Content Protection for Removable Media) into the ATA hard disk specification, we reported on Wednesday. CPRM originates from the the 4C Entity and licensing is administered by License Management International, LLC, which also administers the CSS license.
"This resembles CSS and e-Books: it is another plan to impose additional power over people who use published information, on behalf of those who hope to control the power," he writes in emails to The Register.
"This plan seems to pose a threat to free operating systems. We will surely not be authorized in the US to implement free software to access any of the centrally-controlled data. So a free GNU/Linux system won't be able to do it."
"If users accept the domination of centrally-controlled data, free software faces two dangers, each worse than the other: that users will reject GNU/Linux because it doesn't support the central control over access to these data, or that they will reject free versions of GNU/Linux for versions "enhanced" with proprietary software that support it. Either outcome will be a grave loss for our freedom."
"We must hope that some countries refuse to pass laws to prohibit free software such as DeCSS, so that some part of the world can publish the software that will keep freedom alive, underground, in the rest of the world."
Stallman also highlights the term "copy protection". "The word 'protection'
Indeed: it's a euphemism as incongruous as down-sizing or friendly fire. As an alternative, we quite like "copy control". But if you have snappier suggestions, we'd like to hear them.
--
Nice to hear about the smaller players, especially when they're making inexpensive integrated hardware my mom would like to use.
VIA a small player??? They make chipsets for a good 50% of motherboards (all the AMD ones and quite a few Pentium ones too), they make their own motherboards under the name of FIC, and they make memory too. VIA are anything but small, they're bigger than AMD in terms of capitol if not in terms of profile...
--
For those that want to check out dsniff itself, the URL is:
http://www.monkey.org/~dugsong/dsniff/
Clever stuff...
--
As a parent, I'd much rather have them watch this than Scooby Doo.
What's wrong with Scooby Doo? As Carl Sagan correctly points out in 'A Demon Haunted World' Scooby Doo teaches skepticism. Skepticism is good.
--
As a British subject (remember, we are not citizens), I say we all write to our respective and MPs and request that British Telecom be disallowed to use the word British in its name or any of its literature, for we own this word!
--
Anyway, let me address your points in turn
Has anyone considered the cost of encrypting the majority of Internet traffic? I always encrypt terminal sessions, but the cost of servers encrypting web traffic would be very high. I wouldn't mind; my PC has enough power to encrypt all MY traffic, but how about a busy web server? How much new hardware would slashdot need to support encrypting everything? If slashdot is going to blow that stupid smoke, then slashdot should encrypt all its traffic.
Incase you hadn't noticed hardware companies are producing faster and faster processors all the time, and the price of this increased technology is falling. We are very rapidly approaching (in fact it could be argued were are there) a time in which CPU speed routinely exceeds requirements. Encrypting data produces no overhead in terms of bandwidth, only in terms of CPU time. I'd wager that if the slashdot cru were to go out and replace the hardware they currently have, spending exactly the same amount as they did last time they upgraded, the extra CPU power they'd gain would be enough to add SSL.
Second, is his explicit assumption that Linux is the best thing available, and that free software is always better than proprietary software in quality. Linux does have a lot of good points, and in some cases is the best solution. MS Windows has strengths also. Sometimes Microsoft solutions solve a problem better. Sometimes one of the other systems he ignores is better.
What you say at first is incorrect; the interviewer says Linux is the best thing avaliable, he actually says that free software is the best thing avaliable, which happens to include Linux. The point that is being made is that even if the Microsoft implementation of something were to be technically better (and I'm not going to get into whether or not that is the case), the fact that it is not free makes it inherently worse. Microsoft may make something better, but time does not stand still. If something is free, anyone is free to improve it, and with the potential billions avaliable to improve it, there is no way proprietary software can compete.
--
Mandrake comes with a seriously non standard, patched kernel. (They are actually more than likely using fairly recent emu10k1 drivers, which were probably from opensource.creative.com), they also more than likely patched for usb support and other such little things.
Yup, in fact that should be kernels, it actually comes with two - a 'standard version' which is pretty heavily patched and a 'secure' version with improvements like a non-executable stack and ps showing only the current user's processes. Last time I checked both were still at 2.2.16 (although I'm still using 7.1, don't know if there are more updates for 7.2).
--
Yea, but there's no difference between using a Live! and a cheap soundcard with clean-sounding circitry. Linux doesn't use any of the acceleration features on the sound card, and for the money, there are a lot of cards that have cleaner sound, sans 3D hardware.
I've yet to see a cheap soundcard with anything like as low a S/N ratio as the SB Live! or a DAC even approaching the same quality. I doubt such a beast exists. Of course, unless you've got it hooked up to a decent amp/speakers then you won't hear the differenece.
--
Is that compiler optimisation is probably going to count for more and more as Intel wring every last bit of performance out of x86. Linux distros like Mandrake, which currently comes in a version optimised for P5, could potentially have big performance benefits if you get the version compiled for your specific processor. Of course, the really good news is that Intel et al will have to take quite an active role in GCC development if they want to make their processor look great under free operating systems.
Unfortunately, what it could potentially mean though is that if Intel were to do some sort of special deal with a proprietry OS maker (MS for example) they could make that OS run far faster than any others, simply because it'd be compiled with a better optimised compiler.
-- Piracy is a vicitmless crime, like punching someone in the dark.