This is a bit of a fallacy -- developers don't develop for an ISA, they develop for a platform. </strong><hr></blockquote>
Yes, and the platform is attached to the ISA, 95% of which is attached to the x86 ISA. If it were only a matter of "platforms", we wouldn't be arguing over x86 migration all the time, now would we?
[quote]Yes, and the platform is attached to the ISA, 95% of which is attached to the x86 ISA.<hr></blockquote>
Developers program for the platform and the platform is bound by history to the x86. Because IBM gave their OS contract to Microsoft, which bundled its systems with all cheap clones, etc. Which is why Programmer is right and we're still arguing.
Belive me, the DDR RAM in Power Macs IS being used.</strong><hr></blockquote>
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.
i wouldn't say that quark is in an invincible position. the agency where i work has switched over to indesign 100%. between all the art directors, and studio, that's probably about 40 seats. for such a large, multinational agency, that's a pretty significant switch. people are looking at indesign and quark could be at risk. which is fine with me, i consider quark to be the syquest of the new decade. hahaha.
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.</strong><hr></blockquote>
Wouldn't a 200 MHz FSB be better than a 133 MHz DDR bus?
Wouldn't a 200 MHz FSB be better than a 133 MHz DDR bus?</strong><hr></blockquote>
You would think? Well since no Mac ships with a frontside bus that fast we don't know. But comparing the 1st Gen Dual GHz w/ SDRAM and 133MHz FSB to the 2nd Gen Dual GHz w/ 167MHz FSB and DDR there was little to no differance at all. Anyway..my point in all this was that Apple does not sincerely believe that Quark for X is keeping people from buying Power Macs...they know what the real problems are.
You would think? Well since no Mac ships with a frontside bus that fast we don't know. But comparing the 1st Gen Dual GHz w/ SDRAM and 133MHz FSB to the 2nd Gen Dual GHz w/ 167MHz FSB and DDR there was little to no differance at all. Anyway..my point in all this was that Apple does not sincerely believe that Quark for X is keeping people from buying Power Macs...they know what the real problems are.</strong><hr></blockquote>
I agree with your point.
Just thinking, if Moto does get to a 200 MHz bus with the same DDR set-up that they are using now, that it wouldn't be too bad (along with a 1.5 GHz processor ought to be more than 50% faster than my DP 1GHz QS DA, which is what it will take for me to buy in late January). Not as good as a 900 Mhz bus and 1.8 GHz but enough to hold me over til the 970 or 7457-RM comes out .
FrameMaker _used_ to completely rock. It hasn't been updated practically since Adobe bought Aldus.
The features I use(d) have all been migrated into InDesign. I don't know that I'd expect Adobe to ever Carbonize Frame. Sad but seemingly true.</strong><hr></blockquote>
Cheer up Nevyn!
Pop over to frameusers and search for "OSX": it's made quite clear by various Adobe peeps that OSX compatibility is being worked on. Moreover, there have been reports about Adobe shifting development of Frame8 to India, so it's still clearly a live project.
I'm hoping that they skip over Carbon and go straight for Cocoa. Given Frame's long development cycles, I'm sure that this possibility has been actively considered at Adobe.
PS: Given that Objective C is being talked about in this connection by the right people, we have some reason to hope that this indeed going to be Adobe's approach. F'rinstance:
[quote] FrameMaker on UNIX is X11/Motif based for the UI and dialogs, such that it ships with custom versions of the XLibs and Motif 1.2. All the code would need to be replaced by Aqua-based UI code.
Printing is done by generating an intermediate printer language and running it through fmprintdr.ps to generate PostScript Level 1. That would be replaced by CUPS-aware printing going through the MacOS-X printing model.
There'd be some work involved in converting over the ObjectiveC, but that doesn't seem like a big deal.
Complete replacement of the UI code and printing are the big deals.
...Lee
At 1:16p -0400 10/25/02, Campbell, Art wrote:
>Obviously, your business case is accurate, but it makes me ask what may be a
>naive question: "Why isn't OSX treated as just another flavor of UNIX?" UNIX
>has always been a supported platform so it seems as if porting/compiling to
>support one more flavor would be relatively minor.
<hr></blockquote>
[ 12-28-2002: Message edited by: boy_analog ]
Bugger. That URL didn't work. Falling back to cut & paste....
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.</strong><hr></blockquote>
Please show me a benchmark that is valid - most of those I've seen don't test for I/O.
Our Single 1 GHz Xserve runs JBoss 40% faster than a Dual 1 GHz QuickSilver.
But I guess that your machines aren't on a network, writes to a harddisk, reads from a harddisk, have FireWire or USB hardware connected and so on.
<strong>Yes, and the platform is attached to the ISA, 95% of which is attached to the x86 ISA. If it were only a matter of "platforms", we wouldn't be arguing over x86 migration all the time, now would we?</strong><hr></blockquote>
The x86 migration argument is typically about performance, pricing, availability and the future of processors. It has nothing to do with marketshare (or developer mindshare) except as it results from those points as a secondary effect. Apple switching to x86 will not suddenly result in any additional developer support, in fact a great deal would be lost since all PowerPC code would cease to function (except in emulation where it would be uselessly slow) and developers would feel screwed by Apple again.
The fact that the platform tends to be attached to a particular ISA is irrelevant -- what is relevant is that software is delivered for Win32 or MacOS X regardless of what processor those systems are built atop of. This was specifically in reply to your comment:
[quote]
the rest of the world is running on x86. Development for that ISA will always come first. <hr></blockquote>
A more correct statement would have been:
[quote]
the rest of the world is running on Windows. Development for that platform will always come first.<hr></blockquote>
Apple is a company that builds integrated hardware / software solutions. Its operating system is fundamental to its business and the only way to resolve the development issue that you point out (aside from increasing marketshare to the point where Apple sells most of the world's PCs) would be to abandon its business model in favour of a completely different one, and that would be the end of the Macintosh.
Please show me a benchmark that is valid - most of those I've seen don't test for I/O.
Our Single 1 GHz Xserve runs JBoss 40% faster than a Dual 1 GHz QuickSilver.
But I guess that your machines aren't on a network, writes to a harddisk, reads from a harddisk, have FireWire or USB hardware connected and so on.</strong><hr></blockquote>
<strong>The fact that the platform tends to be attached to a particular ISA is irrelevant -- what is relevant is that software is delivered for Win32 or MacOS X regardless of what processor those systems are built atop of.</strong><hr></blockquote>
Are you trying to separate church from state? Here you say it's irrelevant and before that:
[quote] <strong>
Apple switching to x86 will not suddenly result in any additional developer support, in fact a great deal would be lost since all PowerPC code would cease to function (except in emulation where it would be uselessly slow) and developers would feel screwed by Apple again.</strong><hr></blockquote>
That sounds relevant to me.
I'm looking at the root of the problem, you're looking at the surface. The platform is attached to the ISA. Period. First you say the ISA is irrelevant to the developer, then you say that developers would be screwed by Apple if they switched ISAs. Which is it? I think you're just trying to justify you're agrument that "platform" development doesn't include the ISA, which is BS, and you're intelligent enough to know this.
Apple is a company that builds integrated hardware / software solutions. Its operating system is fundamental to its business and the only way to resolve the development issue that you point out (aside from increasing marketshare to the point where Apple sells most of the world's PCs) would be to abandon its business model in favour of a completely different one, and that would be the end of the Macintosh.
</strong><hr></blockquote>
That wouldn't be the end of Macintosh, it would be the evolution of Macintosh. Think different.
That wouldn't be the end of Macintosh, it would be the evolution of Macintosh. Think different."
Agreed. Part of the Steve Job's 'Cube'/iMac closed box puts many off buying a Mac. £1,395 inc VAT before you can buy a machine you can properly expand and it don't come with a monitor?
Yeesh. PCs are commodity cheapo land now.
Where's an Apple islab for education?
I believe Apple can sell a proprietary x86/Itanium/Sledgehammer m/board.
They'd own that. They own their own software. So they could recompile that into Cocoa.
Fat binaries. With the kind of CPU power that's coming on tap in the next few years? No problem.
Apple need to start thinking beyond PPC if they're going to really convince the other 97% that Apple is 'on board'.
Apple and G5s and Marketshare. There is no single magic wand for Apple. It aint just mhz. It's a bigger problem than it was 18 years ago.
Retail stores. A decent CPU. Perception. Advertising. x86 conundrum. Spec' parity. Price. New, innovative iPod gadjets, best of breed to set Apple apart. The OS.
Sure, Apple have to pick their fights. But most of the obstacles to them getting 10% are within their abilities to control.
A G5 at San Fran would be a start. Why are Mac Quark users going over to PC Quark (alot of cross grades being sold according to Quark!)? Well, price and performance for price are issues. IF Apple doesn't want its print 'bedrock' to be eroded at the same pace as its Edu' 'bedrock' then Apple better move with increasing urgency on the CPU/ram/M'board and Price front.
Must have been alot of Apple 'power'Mac users that gave in and bought PC Towers instead. The 970 might bring 'em back. But only if Apple cuts prices and keeps the dual CPU policy?
And give the 970s a peg to Intel Pentium 4 performance rating? Dual 970? 7.2 GIG PR?
Apple have 4.3 billion in the bank. I can't hlep but feel it's time to use some of it. A Macromedia aquisition? A merger with Adobe?
And I'd keep opening those stores if I was Apple. They need at least another 50.
2003 is going to be another tough year for Apple up to New York. My view.
I'm looking at the root of the problem, you're looking at the surface. The platform is attached to the ISA. Period. First you say the ISA is irrelevant to the developer, then you say that developers would be screwed by Apple if they switched ISAs. Which is it? I think you're just trying to justify you're agrument that "platform" development doesn't include the ISA, which is BS, and you're intelligent enough to know this.</strong><hr></blockquote>
Heh, you can twist it around if you want but my point is sound and isn't a surface issue.
The point is (again) that the platform is the operating system & ISA combination, and Apple switching to x86 does nothing to make the platform more like the WIntel platform because it would still be MacOS X on x86. It would be different than MacOS X on PowerPC, and it would be different from WIntel. Porting MacOS X PowerPC code to MacOS X x86 would be easier than going to WIntel, but it would still require new code releases, more work for developers, testing, etc. Apple has managed to make the Mac more "hardware nimble", but the fact is that there is a great deal of investment in PowerPC right now, and to force another shift upon developers right now might just be enough to kill the platform once and for all. Given plenty of lead time they might be able to make it happen, but it better not be in '03 -- the platform still hasn't stabilized on MacOS X.
One more try: Using the same ISA cannot make two platforms the same, but it makes them different.
I really hate it when somebody brings up "intelligence" during a discussion, its just a cop out. You obviously don't even understand the argument that you're calling BS so you fall back on insults. Come back when you actually understand something about the technology that you're spouting off about.
Comments
<strong>
This is a bit of a fallacy -- developers don't develop for an ISA, they develop for a platform. </strong><hr></blockquote>
Yes, and the platform is attached to the ISA, 95% of which is attached to the x86 ISA. If it were only a matter of "platforms", we wouldn't be arguing over x86 migration all the time, now would we?
[ 12-27-2002: Message edited by: MacLuv ]</p>
Developers program for the platform and the platform is bound by history to the x86. Because IBM gave their OS contract to Microsoft, which bundled its systems with all cheap clones, etc. Which is why Programmer is right and we're still arguing.
<strong>
Belive me, the DDR RAM in Power Macs IS being used.</strong><hr></blockquote>
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.
<strong>
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.</strong><hr></blockquote>
Wouldn't a 200 MHz FSB be better than a 133 MHz DDR bus?
<strong>
Wouldn't a 200 MHz FSB be better than a 133 MHz DDR bus?</strong><hr></blockquote>
You would think? Well since no Mac ships with a frontside bus that fast we don't know. But comparing the 1st Gen Dual GHz w/ SDRAM and 133MHz FSB to the 2nd Gen Dual GHz w/ 167MHz FSB and DDR there was little to no differance at all. Anyway..my point in all this was that Apple does not sincerely believe that Quark for X is keeping people from buying Power Macs...they know what the real problems are.
<strong>
You would think? Well since no Mac ships with a frontside bus that fast we don't know. But comparing the 1st Gen Dual GHz w/ SDRAM and 133MHz FSB to the 2nd Gen Dual GHz w/ 167MHz FSB and DDR there was little to no differance at all. Anyway..my point in all this was that Apple does not sincerely believe that Quark for X is keeping people from buying Power Macs...they know what the real problems are.</strong><hr></blockquote>
I agree with your point.
Just thinking, if Moto does get to a 200 MHz bus with the same DDR set-up that they are using now, that it wouldn't be too bad (along with a 1.5 GHz processor ought to be more than 50% faster than my DP 1GHz QS DA, which is what it will take for me to buy in late January). Not as good as a 900 Mhz bus and 1.8 GHz but enough to hold me over til the 970 or 7457-RM comes out .
<strong>
FrameMaker _used_ to completely rock. It hasn't been updated practically since Adobe bought Aldus.
The features I use(d) have all been migrated into InDesign. I don't know that I'd expect Adobe to ever Carbonize Frame. Sad but seemingly true.</strong><hr></blockquote>
Cheer up Nevyn!
Pop over to frameusers and search for "OSX": it's made quite clear by various Adobe peeps that OSX compatibility is being worked on. Moreover, there have been reports about Adobe shifting development of Frame8 to India, so it's still clearly a live project.
I'm hoping that they skip over Carbon and go straight for Cocoa. Given Frame's long development cycles, I'm sure that this possibility has been actively considered at Adobe.
PS: Given that Objective C is being talked about in this connection by the right people, we have some reason to hope that this indeed going to be Adobe's approach. F'rinstance:
[quote] FrameMaker on UNIX is X11/Motif based for the UI and dialogs, such that it ships with custom versions of the XLibs and Motif 1.2. All the code would need to be replaced by Aqua-based UI code.
Printing is done by generating an intermediate printer language and running it through fmprintdr.ps to generate PostScript Level 1. That would be replaced by CUPS-aware printing going through the MacOS-X printing model.
There'd be some work involved in converting over the ObjectiveC, but that doesn't seem like a big deal.
Complete replacement of the UI code and printing are the big deals.
...Lee
At 1:16p -0400 10/25/02, Campbell, Art wrote:
>Obviously, your business case is accurate, but it makes me ask what may be a
>naive question: "Why isn't OSX treated as just another flavor of UNIX?" UNIX
>has always been a supported platform so it seems as if porting/compiling to
>support one more flavor would be relatively minor.
<hr></blockquote>
[ 12-28-2002: Message edited by: boy_analog ]
Bugger. That URL didn't work. Falling back to cut & paste....
[ 12-28-2002: Message edited by: boy_analog ]
Dang! Hit the wrong button....
[ 12-28-2002: Message edited by: boy_analog ]</p>
<strong>
Missed the point though...It's a perception problem Apple has to deal with. This is fact...300 iMacs instead of Power Macs. Apple would MUCH rather sell Power Macs with a display than an iMac. Problem is that the frontside bus of the G4 does not support ddr ram therefore technically it's a waste of money. Benchmark after benchmark put the 1st Gen Dual Gig right with the 2nd Gen DDR Dual Gig.</strong><hr></blockquote>
Please show me a benchmark that is valid - most of those I've seen don't test for I/O.
Our Single 1 GHz Xserve runs JBoss 40% faster than a Dual 1 GHz QuickSilver.
But I guess that your machines aren't on a network, writes to a harddisk, reads from a harddisk, have FireWire or USB hardware connected and so on.
Hardly surprising.
The 'new' towers weren't that new. They weren't using a brand new processor or quad ram or quad bus.
So it's hardly surprising that any improvement they show might be nominal or at all.
A dual 1.25 costing 2,600 inc Vat pounds? Stinks. Yer can get 3 2.8 Gig Pentium four rigs for that. And I don't like saying it.
Perhaps Apple should do the math and wonder why their 'power'Mac sales are getting the beating of their life.
Lemon Bon Bon
<strong>Yes, and the platform is attached to the ISA, 95% of which is attached to the x86 ISA. If it were only a matter of "platforms", we wouldn't be arguing over x86 migration all the time, now would we?</strong><hr></blockquote>
The x86 migration argument is typically about performance, pricing, availability and the future of processors. It has nothing to do with marketshare (or developer mindshare) except as it results from those points as a secondary effect. Apple switching to x86 will not suddenly result in any additional developer support, in fact a great deal would be lost since all PowerPC code would cease to function (except in emulation where it would be uselessly slow) and developers would feel screwed by Apple again.
The fact that the platform tends to be attached to a particular ISA is irrelevant -- what is relevant is that software is delivered for Win32 or MacOS X regardless of what processor those systems are built atop of. This was specifically in reply to your comment:
[quote]
the rest of the world is running on x86. Development for that ISA will always come first. <hr></blockquote>
A more correct statement would have been:
[quote]
the rest of the world is running on Windows. Development for that platform will always come first.<hr></blockquote>
Apple is a company that builds integrated hardware / software solutions. Its operating system is fundamental to its business and the only way to resolve the development issue that you point out (aside from increasing marketshare to the point where Apple sells most of the world's PCs) would be to abandon its business model in favour of a completely different one, and that would be the end of the Macintosh.
[ 12-28-2002: Message edited by: Programmer ]</p>
<strong>
Please show me a benchmark that is valid - most of those I've seen don't test for I/O.
Our Single 1 GHz Xserve runs JBoss 40% faster than a Dual 1 GHz QuickSilver.
But I guess that your machines aren't on a network, writes to a harddisk, reads from a harddisk, have FireWire or USB hardware connected and so on.</strong><hr></blockquote>
Well here is a <a href="http://arstechnica.infopop.net/OpenTopic/page?a=tpc&s=50009562&f=8300945231&m=6300915245"; target="_blank">thread</a> that began to address the issue. Someone needs to run some XBench tests on the new MDD to see the increases in I/O speeds versus the older machines.
[ 12-28-2002: Message edited by: Bigc ]</p>
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
<strong>The fact that the platform tends to be attached to a particular ISA is irrelevant -- what is relevant is that software is delivered for Win32 or MacOS X regardless of what processor those systems are built atop of.</strong><hr></blockquote>
Are you trying to separate church from state? Here you say it's irrelevant and before that:
[quote] <strong>
Apple switching to x86 will not suddenly result in any additional developer support, in fact a great deal would be lost since all PowerPC code would cease to function (except in emulation where it would be uselessly slow) and developers would feel screwed by Apple again.</strong><hr></blockquote>
That sounds relevant to me.
I'm looking at the root of the problem, you're looking at the surface. The platform is attached to the ISA. Period. First you say the ISA is irrelevant to the developer, then you say that developers would be screwed by Apple if they switched ISAs. Which is it? I think you're just trying to justify you're agrument that "platform" development doesn't include the ISA, which is BS, and you're intelligent enough to know this.
<strong>
Apple is a company that builds integrated hardware / software solutions. Its operating system is fundamental to its business and the only way to resolve the development issue that you point out (aside from increasing marketshare to the point where Apple sells most of the world's PCs) would be to abandon its business model in favour of a completely different one, and that would be the end of the Macintosh.
</strong><hr></blockquote>
That wouldn't be the end of Macintosh, it would be the evolution of Macintosh. Think different.
That wouldn't be the end of Macintosh, it would be the evolution of Macintosh. Think different."
Agreed. Part of the Steve Job's 'Cube'/iMac closed box puts many off buying a Mac. £1,395 inc VAT before you can buy a machine you can properly expand and it don't come with a monitor?
Yeesh. PCs are commodity cheapo land now.
Where's an Apple islab for education?
I believe Apple can sell a proprietary x86/Itanium/Sledgehammer m/board.
They'd own that. They own their own software. So they could recompile that into Cocoa.
Fat binaries. With the kind of CPU power that's coming on tap in the next few years? No problem.
Apple need to start thinking beyond PPC if they're going to really convince the other 97% that Apple is 'on board'.
Apple and G5s and Marketshare. There is no single magic wand for Apple. It aint just mhz. It's a bigger problem than it was 18 years ago.
Retail stores. A decent CPU. Perception. Advertising. x86 conundrum. Spec' parity. Price. New, innovative iPod gadjets, best of breed to set Apple apart. The OS.
Sure, Apple have to pick their fights. But most of the obstacles to them getting 10% are within their abilities to control.
A G5 at San Fran would be a start. Why are Mac Quark users going over to PC Quark (alot of cross grades being sold according to Quark!)? Well, price and performance for price are issues. IF Apple doesn't want its print 'bedrock' to be eroded at the same pace as its Edu' 'bedrock' then Apple better move with increasing urgency on the CPU/ram/M'board and Price front.
Must have been alot of Apple 'power'Mac users that gave in and bought PC Towers instead. The 970 might bring 'em back. But only if Apple cuts prices and keeps the dual CPU policy?
And give the 970s a peg to Intel Pentium 4 performance rating? Dual 970? 7.2 GIG PR?
Apple have 4.3 billion in the bank. I can't hlep but feel it's time to use some of it. A Macromedia aquisition? A merger with Adobe?
And I'd keep opening those stores if I was Apple. They need at least another 50.
2003 is going to be another tough year for Apple up to New York. My view.
Lemon Bon Bon
[ 12-28-2002: Message edited by: Lemon Bon Bon ]</p>
[ 12-28-2002: Message edited by: Barto ]</p>
<strong>
That sounds relevant to me.
I'm looking at the root of the problem, you're looking at the surface. The platform is attached to the ISA. Period. First you say the ISA is irrelevant to the developer, then you say that developers would be screwed by Apple if they switched ISAs. Which is it? I think you're just trying to justify you're agrument that "platform" development doesn't include the ISA, which is BS, and you're intelligent enough to know this.</strong><hr></blockquote>
Heh, you can twist it around if you want but my point is sound and isn't a surface issue.
The point is (again) that the platform is the operating system & ISA combination, and Apple switching to x86 does nothing to make the platform more like the WIntel platform because it would still be MacOS X on x86. It would be different than MacOS X on PowerPC, and it would be different from WIntel. Porting MacOS X PowerPC code to MacOS X x86 would be easier than going to WIntel, but it would still require new code releases, more work for developers, testing, etc. Apple has managed to make the Mac more "hardware nimble", but the fact is that there is a great deal of investment in PowerPC right now, and to force another shift upon developers right now might just be enough to kill the platform once and for all. Given plenty of lead time they might be able to make it happen, but it better not be in '03 -- the platform still hasn't stabilized on MacOS X.
One more try: Using the same ISA cannot make two platforms the same, but it makes them different.
I really hate it when somebody brings up "intelligence" during a discussion, its just a cop out. You obviously don't even understand the argument that you're calling BS so you fall back on insults. Come back when you actually understand something about the technology that you're spouting off about.