FROM CLIVE ------What are you on!? First you're saying 970s for MWSF, now you're talking Q2 03, which in a worst case is going to be close to six months later.
You have no logic because you have no understanding of what "sampling" means - test units, in small quantities, not for production.
-------------------------------------------
I know what sampling mean and I read IBM press release. Thank you very much. But its just a press release, this release is for general public as well as other smaller computer customers. Apple is probably the most important one because SJ is talking about market share and who wouldn't like it. Semiconductor firms like volume. If importance is a key, anything is possible.
It's safe to completely, totally, 100% assume that the "G5" (at least, as we would know it) is dead, right? I've followed AI as well as the stories from Architosh as well as other sites and it's my general understanding that for all intents and purposes, you can pretty much stick a fork in the G5...
I'd like to know from the more technically-minded people here just how possible it is for these reputed new versions of the G4 (7457, 7457-RM) to actually be produced. From what I understand, trying to graft on a faster bus or DDR-support onto the G4 would be almost impossible...</strong><hr></blockquote>
There is no evidence that a Motorola G5 desktop processor exists.
The 7455 currently shipping has a slightly faster bus (167 MHz) and the rumours for the 7457 claim faster still (200 MHz). The leaked Moto documents which refer to the 7457-RM describe a G4 variant with an on-chip memory controller and a RapidIO bus. The 7457-RM would be a fairly significant project at Moto, but far less effort than a fully new core. We aren't likely to see the 7457-RM for quite a while, if ever. Hopefully the 7457 ships Real Soon Now.
<strong>That's the thing, IBM uses both fiscal and real Q-dates. It's rare that they distinguish, so for all we know it may be real or fiscal. Optimism says it's fiscal, but reality says they are talking about a real Q2.</strong><hr></blockquote>
The 970 was announced mid-October, and they said sampling Q1 '03, production 2H '03. That's less than a 1 quarter delay from the announcement. If it was earlier then they probably would have said "sampling this quarter" or "sampling now". They also said production in second half which isn't the same as Q3.
I think way too much attention is being paid to Apple not "allowing" IBM to announce anything. Before the recent series of G4s most of the PowerPC chips were announced at the October conference well in advance of Apple's adopting them in product. And that was when IBM was involved.
Anecdotal evidence about MacOS X adoption is pretty much pointless as we have no way to judge who is a more representative sample. Since Jaguar arrived, however, MacOS X has become a much more usable system. I bought a new Mac to run it, and I know a few other "professionals" who bought new machines and they are all running Jaguar. One of them is also single-handedly responsible for evangelizing at least two of shops into converting from Win32 machines to MacOS X. There are a few areas where adoption is being held up by software (audio & Quark being the well known ones), but for a large number of users Jaguar is just fine. Terrific, actually. I'm very happy with the move from MacOS 9.1 to 10.2 -- I haven't yet booted into MacOS 9, and aim to never do so. I rarely even launch Classic. For an immature OS it feels remarkably solid compared to the Win2K environment I have to work in all day at the office. Yes this is all anecdotal too, but I suspect that many users have straightforward needs that are being met very well by X.
<strong> The 7457-RM would be a fairly significant project at Moto, but far less effort than a fully new core. We aren't likely to see the 7457-RM for quite a while, if ever.</strong><hr></blockquote>
Ah, that was kind of what I was trying to figure out. Thanks for the info. What I actually was trying to find out was what the technical barriers were in the G4 that makes DDR support or a faster bus so darned hard to put into the G4...I don't suppose you could shed some light on this?
<strong>Ah, that was kind of what I was trying to figure out. Thanks for the info. What I actually was trying to find out was what the technical barriers were in the G4 that makes DDR support or a faster bus so darned hard to put into the G4...I don't suppose you could shed some light on this?</strong><hr></blockquote>
Uh, its a lot of work?
Heh, sorry for being cheeky. I can't really tell you why because I'm a software guy, not a hardware guy. Even if I could tell you in detail you probably wouldn't understand the description and it wouldn't fit in this little message box. In cases like this I usually fall back on an analogy... imagine Porsche taking its design for the Boxster-S and trying to convert it into a tracked hill climber. They could do it, even use some parts that were already designed for something else, but it would be quite a bit of work to build something marketable.
[quote]Originally posted by Transcendental Octothorpe:
<strong>
That might just be their only option.
I hope as much as anyone that the 970 will appear this Jan, but I don't really expect it. Apple may not have been able to swing a deal with IBM that required IBM's total silence on the 970 until Apple's release. Bear in mind that Apple didn't really have any other options, so if IBM says "we don't work that way, take it or leave it" Apple had little choice.
Oh, and for those still unsure whether Apple is even going to USE the 970 (some still seem to doubt), I have no doubt. I have been told by those who would know at IBM that "we have won all Apple business". Think what you like.</strong><hr></blockquote>
IBM's total not be good for Apple. Apple needs IBM to spend as much money as possable telling the world how great the 970 is, so that when Apple finally anounces that they are shipping 970 based computers they dont have to sell the 970's merits as well as their computer. What Apple dosnt need IBM to do is say, "oh yea, next month Apple is releasing a Quad 970 based computer for under $5000."
I believe the SEC rules governing the release of information require a company to use calendar dates. If IBM said 2H then they meant after June. Also, IBM can't mislead through public disclosure. They can't say 2H and then the 970 show up in January. Everyone here could think of a fruity stock that would be substantially affected were that to happen.
Heh, sorry for being cheeky. I can't really tell you why because I'm a software guy, not a hardware guy. Even if I could tell you in detail you probably wouldn't understand the description and it wouldn't fit in this little message box. In cases like this I usually fall back on an analogy... imagine Porsche taking its design for the Boxster-S and trying to convert it into a tracked hill climber. They could do it, even use some parts that were already designed for something else, but it would be quite a bit of work to build something marketable.
Does that help? </strong><hr></blockquote>
Not meaning to nitpick, but Porsche has most of those parts develready for the Carra 4 and the 911 before it, which various versions have run the Baha 500 for many years (at least back to the 80s). The Boxter uses a lot of standard 911 components, the biggest problem that they would have to overcome would be how to raise the Boxter so that it could have the needed ground clearence. Even then it would probably be a very stable performer due to the very low center of gravety and weight distribution thanks to the flat mid-engine design that goes back to the 914, 917, 904, and the 550.
<strong>Not meaning to nitpick, but Porsche has most of those parts develready for the Carra 4 and the 911 before it, which various versions have run the Baha 500 for many years (at least back to the 80s). The Boxter uses a lot of standard 911 components, the biggest problem that they would have to overcome would be how to raise the Boxter so that it could have the needed ground clearence. Even then it would probably be a very stable performer due to the very low center of gravety and weight distribution thanks to the flat mid-engine design that goes back to the 914, 917, 904, and the 550.</strong><hr></blockquote>
That's not nitpicking, that's exactly right. And Motorola has a SoC fabric, memory controller & RIO interface in the PPC 8xxx which they could retrofit in place of the G4's bus interface and load/store units. It would still be a piece of work.
PPC970 aka PowerMac G5 will debut (at least be announced) at MWSF, with double pumped FSB (low end = 266, high end = 333.) that DDR ram can finally flex it's muscles.
Considering the manners of the poster, are we sure that Macluv didnt start this thread?
Just a theory....
Anyway I just have to refute this.
[quote] Mac platform lacks apps and file formats are not compatible <hr></blockquote>
Have you ever heard of Quicktime, Graphic Converter and Debabelizer? They have been with the platform for years and are on OSX now. GC is shareware for cry eye!
An OSX machine out of the box supports more file formats than any PC with Windows or Linux. Windows supports primarily bmp for graphics and AVI for video PERIOD. If not for third party developers that would be just about it even if we weren't talking about a machine just out of the box.
However, fine I'll bite on this thread. I agree that IBM is probably talking Calendar rather than fiscal. No way in January and believe me I wish I was wrong.
That's because educational customers CAN order which OS the want on the machine.</strong><hr></blockquote>
And they all order directly from Apple?
Either way, come up with some method to derive these figures and I'll roll over - otherwise it's just PR hype.
Just to re-emphasise the figures: if Apple says 80% of our professional customers, we can guess that they mean some subset of PowerMac buyers - Apple's own figures say that PowerMac sales were only 25% of units sold last year, therefore this huge figure actually means they are referring to just 20% of new machines sold. Chances are they are actually referring to sales in the last quarter, or at best in the last half of the year - so the "professional user" they are citing probably accounts for somewhere between 5-10% of units sold last year.
Not such a good headline though, is it.
Come up with your own construction, if you like. That's what Apple did.
<strong>1. My quoted sample is small, but that was just last week's work. I can say with a degree of certainty that out of approximately 120 Macs across four sites precisely two are running X on a fulltime basis.</strong><hr></blockquote>
Sorry, that's a small lie. I just remembered that there's an additional G3 iMac that is running X (plus two G4 towers), but AFAIK it was bought for some unknown reason and no one ever uses it - it's always in sleep mode when I see it and has no apps installed.
And they all order directly from Apple?</strong><hr></blockquote>
We're talking educational institutions here - not single students.
[quote]<strong>Just to re-emphasise the figures: if Apple says 80% of our professional customers, we can guess that they mean some subset of PowerMac buyers - Apple's own figures say that PowerMac sales were only 25% of units sold last year, therefore this huge figure actually means they are referring to just 20% of new machines sold.</strong><hr></blockquote>
That does not mean that the rest of the Macs bought are running Mac OS 9.
Furthermore PowerBooks would fit into the pro segment.
[quote]<strong>Chances are they are actually referring to sales in the last quarter, or at best in the last half of the year - so the "professional user" they are citing probably accounts for somewhere between 5-10% of units sold last year.</strong><hr></blockquote>
They are talking about the current status. When you look at sales to decide your strategy, you use the sale of previous quarters to get an idea of what the trend is, and currently 80% of all professional buyers are using Mac OS X.
Not all of Apple's pro customers are in the publishing segment.
The number for non-professionals are probably higher.
Sorry, that's a small lie. I just remembered that there's an additional G3 iMac that is running X (plus two G4 towers), but AFAIK it was bought for some unknown reason and no one ever uses it - it's always in sleep mode when I see it and has no apps installed.</strong><hr></blockquote>
Gee, no apps! And you wonder why noone use it?
Lots of graphic pros have made the switch, and when you switch, you switch.
You just don't place a Mac in the corner that people can 'play' with - most tend to work at their computer.
That does not mean that the rest of the Macs bought are running Mac OS 9.
Furthermore PowerBooks would fit into the pro segment.
</strong><hr></blockquote>
Furthermore, whogivesafsck? You have no evidence except what you see, I have no evidence except what I see - but I think that we can both understand that Apple's 80% figure is pulled out of their arse.
We all know that Apple is going to struggle to get to their stated target of 20% of the *entire* userbase running X by the end of 2002 - for a lot of reasons: sales of new hardware are down, inertia on the part of users, Quark not native (this is a red herring)... So, coming out with figures stating 50% of this, or 80% of that (ie multiple times the stated goal) are simply meaningless - to argue otherwise is just stupidity.
Perhaps I'm not being clear enough? Someone, not me, bought the iMac. It's unclear why they bought it, and no one seems to have a use for it. So as to state its obvious uselessness "it has no apps on it". To further state its uselessness in context - no one could be bothered to even install 9.x/delete X/set it to start-up in 9.x - they just left it pretty much how it came out of the box (hint, the dice is loaded in X's favour as the "default" OS).
That this machine is running X in at this site is a symbol of its redundancy, nothing else.
[quote]Originally posted by JLL:
<strong>
Lots of graphic pros have made the switch, and when you switch, you switch.</strong><hr></blockquote>
Well, I know quite a lot of graphics pros, all over the world, and most of them are using 9.x.
As far as "when you switch, you switch" goes - what's that mean? You're so enamoured with X that you cant stand 9.x anymore? If that's what you think then you couldn't be more wrong. I know plenty of "early adoptors" who've "tried" to switch - and regularly "re-try" to switch, but end up bouncing back to 9.x because they just want to get some work done.
I've no idea what you do, but I work in graphics, prepress and new media - I've got around 16 years commercial experience, and I've been using Macs since 1988. As well as being a "production" animal my company supports three other commercial studios, totalling around 120 Macs (and perhaps 15-20 Windows boxes, plus a couple of Suns). I have friends working in large studios in many countries within the EU, US and Australia - so I think I have a pretty good grasp on what "graphic pros" are doing: trying to spend as much time as possible in a productive environment, with as few hitches as possible.
You know what? 9.x gives them that environment, in known and reliable quantities. X does not.
You'll note that this argument is analagous to the XPress v InDesign debate.
Comments
You have no logic because you have no understanding of what "sampling" means - test units, in small quantities, not for production.
-------------------------------------------
I know what sampling mean and I read IBM press release. Thank you very much. But its just a press release, this release is for general public as well as other smaller computer customers. Apple is probably the most important one because SJ is talking about market share and who wouldn't like it. Semiconductor firms like volume. If importance is a key, anything is possible.
<strong>Trying to go back on topic...
It's safe to completely, totally, 100% assume that the "G5" (at least, as we would know it) is dead, right? I've followed AI as well as the stories from Architosh as well as other sites and it's my general understanding that for all intents and purposes, you can pretty much stick a fork in the G5...
I'd like to know from the more technically-minded people here just how possible it is for these reputed new versions of the G4 (7457, 7457-RM) to actually be produced. From what I understand, trying to graft on a faster bus or DDR-support onto the G4 would be almost impossible...</strong><hr></blockquote>
There is no evidence that a Motorola G5 desktop processor exists.
The 7455 currently shipping has a slightly faster bus (167 MHz) and the rumours for the 7457 claim faster still (200 MHz). The leaked Moto documents which refer to the 7457-RM describe a G4 variant with an on-chip memory controller and a RapidIO bus. The 7457-RM would be a fairly significant project at Moto, but far less effort than a fully new core. We aren't likely to see the 7457-RM for quite a while, if ever. Hopefully the 7457 ships Real Soon Now.
<strong>That's the thing, IBM uses both fiscal and real Q-dates. It's rare that they distinguish, so for all we know it may be real or fiscal. Optimism says it's fiscal, but reality says they are talking about a real Q2.</strong><hr></blockquote>
The 970 was announced mid-October, and they said sampling Q1 '03, production 2H '03. That's less than a 1 quarter delay from the announcement. If it was earlier then they probably would have said "sampling this quarter" or "sampling now". They also said production in second half which isn't the same as Q3.
I think way too much attention is being paid to Apple not "allowing" IBM to announce anything. Before the recent series of G4s most of the PowerPC chips were announced at the October conference well in advance of Apple's adopting them in product. And that was when IBM was involved.
Anecdotal evidence about MacOS X adoption is pretty much pointless as we have no way to judge who is a more representative sample. Since Jaguar arrived, however, MacOS X has become a much more usable system. I bought a new Mac to run it, and I know a few other "professionals" who bought new machines and they are all running Jaguar. One of them is also single-handedly responsible for evangelizing at least two of shops into converting from Win32 machines to MacOS X. There are a few areas where adoption is being held up by software (audio & Quark being the well known ones), but for a large number of users Jaguar is just fine. Terrific, actually. I'm very happy with the move from MacOS 9.1 to 10.2 -- I haven't yet booted into MacOS 9, and aim to never do so. I rarely even launch Classic. For an immature OS it feels remarkably solid compared to the Win2K environment I have to work in all day at the office. Yes this is all anecdotal too, but I suspect that many users have straightforward needs that are being met very well by X.
<strong> The 7457-RM would be a fairly significant project at Moto, but far less effort than a fully new core. We aren't likely to see the 7457-RM for quite a while, if ever.</strong><hr></blockquote>
Ah, that was kind of what I was trying to figure out. Thanks for the info. What I actually was trying to find out was what the technical barriers were in the G4 that makes DDR support or a faster bus so darned hard to put into the G4...I don't suppose you could shed some light on this?
<strong>Ah, that was kind of what I was trying to figure out. Thanks for the info. What I actually was trying to find out was what the technical barriers were in the G4 that makes DDR support or a faster bus so darned hard to put into the G4...I don't suppose you could shed some light on this?</strong><hr></blockquote>
Uh, its a lot of work?
Heh, sorry for being cheeky. I can't really tell you why because I'm a software guy, not a hardware guy. Even if I could tell you in detail you probably wouldn't understand the description and it wouldn't fit in this little message box. In cases like this I usually fall back on an analogy... imagine Porsche taking its design for the Boxster-S and trying to convert it into a tracked hill climber. They could do it, even use some parts that were already designed for something else, but it would be quite a bit of work to build something marketable.
Does that help?
<strong>
That might just be their only option.
I hope as much as anyone that the 970 will appear this Jan, but I don't really expect it. Apple may not have been able to swing a deal with IBM that required IBM's total silence on the 970 until Apple's release. Bear in mind that Apple didn't really have any other options, so if IBM says "we don't work that way, take it or leave it" Apple had little choice.
Oh, and for those still unsure whether Apple is even going to USE the 970 (some still seem to doubt), I have no doubt. I have been told by those who would know at IBM that "we have won all Apple business". Think what you like.</strong><hr></blockquote>
IBM's total not be good for Apple. Apple needs IBM to spend as much money as possable telling the world how great the 970 is, so that when Apple finally anounces that they are shipping 970 based computers they dont have to sell the 970's merits as well as their computer. What Apple dosnt need IBM to do is say, "oh yea, next month Apple is releasing a Quad 970 based computer for under $5000."
BTW, I hate Monobasic Sodium Phosphate!!
<strong>
Uh, its a lot of work?
Heh, sorry for being cheeky. I can't really tell you why because I'm a software guy, not a hardware guy. Even if I could tell you in detail you probably wouldn't understand the description and it wouldn't fit in this little message box. In cases like this I usually fall back on an analogy... imagine Porsche taking its design for the Boxster-S and trying to convert it into a tracked hill climber. They could do it, even use some parts that were already designed for something else, but it would be quite a bit of work to build something marketable.
Does that help?
Not meaning to nitpick, but Porsche has most of those parts develready for the Carra 4 and the 911 before it, which various versions have run the Baha 500 for many years (at least back to the 80s). The Boxter uses a lot of standard 911 components, the biggest problem that they would have to overcome would be how to raise the Boxter so that it could have the needed ground clearence. Even then it would probably be a very stable performer due to the very low center of gravety and weight distribution thanks to the flat mid-engine design that goes back to the 914, 917, 904, and the 550.
<strong>Not meaning to nitpick, but Porsche has most of those parts develready for the Carra 4 and the 911 before it, which various versions have run the Baha 500 for many years (at least back to the 80s). The Boxter uses a lot of standard 911 components, the biggest problem that they would have to overcome would be how to raise the Boxter so that it could have the needed ground clearence. Even then it would probably be a very stable performer due to the very low center of gravety and weight distribution thanks to the flat mid-engine design that goes back to the 914, 917, 904, and the 550.</strong><hr></blockquote>
That's not nitpicking, that's exactly right. And Motorola has a SoC fabric, memory controller & RIO interface in the PPC 8xxx which they could retrofit in place of the G4's bus interface and load/store units. It would still be a piece of work.
New enclosure ? hell yeah glass-metallic.
<strong>That'll make all the difference - PR blather, that's all that is. Either way:
"The company said that educational customers are ordering 50% of the its Macs with Mac OS X as the default OS"</strong><hr></blockquote>
That's because educational customers CAN order which OS the want on the machine.
Just a theory....
Anyway I just have to refute this.
[quote] Mac platform lacks apps and file formats are not compatible <hr></blockquote>
Have you ever heard of Quicktime, Graphic Converter and Debabelizer? They have been with the platform for years and are on OSX now. GC is shareware for cry eye!
An OSX machine out of the box supports more file formats than any PC with Windows or Linux. Windows supports primarily bmp for graphics and AVI for video PERIOD. If not for third party developers that would be just about it even if we weren't talking about a machine just out of the box.
However, fine I'll bite on this thread. I agree that IBM is probably talking Calendar rather than fiscal. No way in January and believe me I wish I was wrong.
<strong>
That's because educational customers CAN order which OS the want on the machine.</strong><hr></blockquote>
And they all order directly from Apple?
Either way, come up with some method to derive these figures and I'll roll over - otherwise it's just PR hype.
Just to re-emphasise the figures: if Apple says 80% of our professional customers, we can guess that they mean some subset of PowerMac buyers - Apple's own figures say that PowerMac sales were only 25% of units sold last year, therefore this huge figure actually means they are referring to just 20% of new machines sold. Chances are they are actually referring to sales in the last quarter, or at best in the last half of the year - so the "professional user" they are citing probably accounts for somewhere between 5-10% of units sold last year.
Not such a good headline though, is it.
Come up with your own construction, if you like. That's what Apple did.
<strong>1. My quoted sample is small, but that was just last week's work. I can say with a degree of certainty that out of approximately 120 Macs across four sites precisely two are running X on a fulltime basis.</strong><hr></blockquote>
Sorry, that's a small lie. I just remembered that there's an additional G3 iMac that is running X (plus two G4 towers), but AFAIK it was bought for some unknown reason and no one ever uses it - it's always in sleep mode when I see it and has no apps installed.
<strong>
And they all order directly from Apple?</strong><hr></blockquote>
We're talking educational institutions here - not single students.
[quote]<strong>Just to re-emphasise the figures: if Apple says 80% of our professional customers, we can guess that they mean some subset of PowerMac buyers - Apple's own figures say that PowerMac sales were only 25% of units sold last year, therefore this huge figure actually means they are referring to just 20% of new machines sold.</strong><hr></blockquote>
That does not mean that the rest of the Macs bought are running Mac OS 9.
Furthermore PowerBooks would fit into the pro segment.
[quote]<strong>Chances are they are actually referring to sales in the last quarter, or at best in the last half of the year - so the "professional user" they are citing probably accounts for somewhere between 5-10% of units sold last year.</strong><hr></blockquote>
They are talking about the current status. When you look at sales to decide your strategy, you use the sale of previous quarters to get an idea of what the trend is, and currently 80% of all professional buyers are using Mac OS X.
Not all of Apple's pro customers are in the publishing segment.
The number for non-professionals are probably higher.
<strong>
Sorry, that's a small lie. I just remembered that there's an additional G3 iMac that is running X (plus two G4 towers), but AFAIK it was bought for some unknown reason and no one ever uses it - it's always in sleep mode when I see it and has no apps installed.</strong><hr></blockquote>
Gee, no apps! And you wonder why noone use it?
Lots of graphic pros have made the switch, and when you switch, you switch.
You just don't place a Mac in the corner that people can 'play' with - most tend to work at their computer.
<strong>
That does not mean that the rest of the Macs bought are running Mac OS 9.
Furthermore PowerBooks would fit into the pro segment.
</strong><hr></blockquote>
Furthermore, whogivesafsck? You have no evidence except what you see, I have no evidence except what I see - but I think that we can both understand that Apple's 80% figure is pulled out of their arse.
We all know that Apple is going to struggle to get to their stated target of 20% of the *entire* userbase running X by the end of 2002 - for a lot of reasons: sales of new hardware are down, inertia on the part of users, Quark not native (this is a red herring)... So, coming out with figures stating 50% of this, or 80% of that (ie multiple times the stated goal) are simply meaningless - to argue otherwise is just stupidity.
<strong>
Gee, no apps! And you wonder why noone use it?
</strong><hr></blockquote>
Perhaps I'm not being clear enough? Someone, not me, bought the iMac. It's unclear why they bought it, and no one seems to have a use for it. So as to state its obvious uselessness "it has no apps on it". To further state its uselessness in context - no one could be bothered to even install 9.x/delete X/set it to start-up in 9.x - they just left it pretty much how it came out of the box (hint, the dice is loaded in X's favour as the "default" OS).
That this machine is running X in at this site is a symbol of its redundancy, nothing else.
[quote]Originally posted by JLL:
<strong>
Lots of graphic pros have made the switch, and when you switch, you switch.</strong><hr></blockquote>
Well, I know quite a lot of graphics pros, all over the world, and most of them are using 9.x.
As far as "when you switch, you switch" goes - what's that mean? You're so enamoured with X that you cant stand 9.x anymore? If that's what you think then you couldn't be more wrong. I know plenty of "early adoptors" who've "tried" to switch - and regularly "re-try" to switch, but end up bouncing back to 9.x because they just want to get some work done.
I've no idea what you do, but I work in graphics, prepress and new media - I've got around 16 years commercial experience, and I've been using Macs since 1988. As well as being a "production" animal my company supports three other commercial studios, totalling around 120 Macs (and perhaps 15-20 Windows boxes, plus a couple of Suns). I have friends working in large studios in many countries within the EU, US and Australia - so I think I have a pretty good grasp on what "graphic pros" are doing: trying to spend as much time as possible in a productive environment, with as few hitches as possible.
You know what? 9.x gives them that environment, in known and reliable quantities. X does not.
You'll note that this argument is analagous to the XPress v InDesign debate.
<strong>...but I think that we can both understand that Apple's 80% figure is pulled out of their arse.</strong><hr></blockquote>
Actually they probably get that figure from things like .Mac connection information and tech support calls.