?Byer in his official blog pointed out that many of the perceived benefits of 64-bit computing simply won't manifest themselves with current generations of hardware and software.?
Perhaps, some of you should take a few moments and read Byer's official blog. It seems the comments and his responses are more educated and informative that what is appearing here.
The program doesn't have to be 64 bit. That's a misconception. It does need to address more than 32 bits of memory. That isn't too difficult to impliment. They can do that later. The chipsets themselves don't address 64 bits for RAM anyway.
Most filters are already being handled by either Altivec on the PPC, or SSE on Intel. Both address 128 bits.
We don't need 64 bit menus and such.
Hmm. If it was that trivial to implement 64bit functionality into Photoshop then surely they would have done so already?
I know there is a plug-in which passes certain calculations through Altivec and there is an equivalent MMX/SSE plug-in for Windows. When the G5 came out it came with an updated version of the Altivec plug-in which was supposed to "take advantage of the power of the G5" of some such nonsense, but as far as I know all it did was add support for IBM's version of Altivec (VMX, or whatever it's called).
However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM.
The difficulty for Adobe is to update the way the core image, layer and blending code works. Not to mention their entire virtual memory management system. 64bit systems have been out for a while now, and I just find it irritating that despite the constant request from customers for the last 4 years to allow Photoshop to access more memory Adobe is brushing them off and giving us "new" features in CS3 which are either ripped from Photoshop Elements or Illustrator CS.
...However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM....
I agree. But how many pros can afford say 8gb-10gb Mac Pro systems? Intel Core2 [Conroe] tops out at 4gb. It would be a fairly elite group that has the RAM to go past that 3gb limit.
However, as I mentioned in my projections of RAM spec's becoming "standard" as the years go by, in 2009 probably is when we're looking at dumping the whole project into 10gbs of your 16+gb machine. CS4....
AS others have indicated it is difficult to give value to the logic of his argument when he is talking about an operating system that will be replaced by something that totally devalues his argument by the time his product comes out. Makes him sound like an Estate Agent. Worse still one can't but feel that om past experience and prsent promises a 64 bit version of photoship won't be around for a further year or so after CS3 launch when the equipment and software to take advantage of it will be quite mature to say the least. Fine if you want to be a follower but if Adobe don't want to lead the moves to superior computing at least be honest about it and accept that this is about money over innovation. Sounds very Microsoft to me which probably shows how they vewiw the world of competition and develop accordingly. A shame but a serious opportunity for others to compete with what Adobe think are dominant products.
Let Apple just buy Adobe and squash this whole matter! Why shouldn't they? If one company like Adobe is going to have that much sway on the sales and development of Apple's current and future products, wouldn't Apple benefit greatly is using their massive war chest to buy out Chizen's little monopoly?
I fully expect Apple to come out with another app that chips away at Adobe's capabilities. They did it Final Cut, iPages, iWeb, Apeture, and Preview/PDF. So a more advanced image editing app than iPhoto and Apeture would seem natural (especially w/ CoreImage).
While that's all well and fine, I agree w/ an earlier post that Adobe's lineup is more universally accepted if not absolutely required for each field. Instead of developing a wannabe, Apple is better off taking over Adobe and controlling the farm.
The Acrobat line purchase alone would scare the heck out of Microsoft!
I expect the real reason they don't want to add 64bit support is that it is going to be a major rewrite of the code and it'll be a nightmare to make it work on both Windows and Mac OS X.
Cross-platform compatibility and grotesquely old legacy code are the main things which are stopping development of Photoshop these days.
I agree with these comments. I'm sure Adobe's plate is full just transitioning to Mac Intel. To undertake a major rewrite of the very foundations of Photoshop would be too much to manage at the same time. And then there's that cross platform concern--about which I care nada, but Adobe does.
So, they'll rush out and buy CS3, but they'll leave Leopard on the shelf? OK, we get it. The smarties are telling us why 64-bit doesn't matter... yet. I get that. The point, though, is that if they were really pushing the envelope and not just cashing in on incremental updates, they would be there IN ADVANCE. It will matter.
Plus, "Pros and shops" don't act as a group. Some are early adopters and have the Mac Pros that can handle it, some are still using clusters of G5s that they invested so much in a while ago.
My understanding and experience is that people that do paying creative work with their computers are usually at least slightly conservative with their software, because of their past experiences of rushing into new software. They might buy the software but that doesn't mean that they will use it right away. For me, the released version of Tiger wasn't all that great until after a few updates, and I don't expect Leopard to be either. The pros will probably buy CS3 but will all work switch to it right away? Probably not except for the daring. The rest might experiment with it from time to time or try it for less time-critical projects to get a feel for its stability and acclimate to the changes.
Quote:
P.S. What Apple does with FREE programs is irrelevant.
JeffDM - "The filters wouldn't benefit from 64 bit code? I suppose that's all handled by vector ops now. But I thought people were begging Adobe to have something to handle images larger than 2GB for many years." You are right...
It's possible for a 32 bit program to handle larger images. The GIMP has done this for several years by transparently breaking up the working data into "tiles". When I remembered this, I realized that there is not a sufficient reason to port it to 64 bit yet.
I'm surprised how nobody has mentioned the tools that Adobe ripped off from Apple in their beta of CS3. Like the loop? That's a blatant rip-off if I've ever seen one. I wouldn't be surprised if there was a lawsuit soon.
Happy holidays to you over there across the world!
This (multiple cpu's), is actually far more important these days.
Adobe has always said that PS worked with multiple cpu's.
And they were right.
But, they didn't imagine that the word multiple would mean more than two!
So now they are scrambling to increase the number of threads. Not easy, but not impossible for the work the program is called to do.
Properly handled, it will work well with 3 to 4 GB's of RAM.
Someone here said that the program called for 7GB of RAM when rotating an image or some such operation. I can't imagine why it would. I've not been able to verify that on a smaller scale.
PS has always, according to Adobe, needed at least 3 times as much RAM as the largest image being worked on at the time for 100% efficiency, that is, all of the work being done in RAM, and as much as 5 times under extreme circumstances. The efficiency report, always available from the bottom of the current window, will let one know.
Working with many cores actually allows less memory to be needed at any one time, as the image is broken into smaller boxes. So there are other beneficial effects from going multiple core.
Of course, all this is mitigated by what plug-ins one has running at the time, and whether or not multiple images are open for use in comps.
But, rarely are people working on multiple 300 MB images at one time as we use to do in my place.
Even when we worked on 500 MB+ images for large billboards, the question of HD speeds rarely entered into the discussion. A fast RAID took care of most of that, and the delays were hardly ever more than a few seconds.
Today, drives, and the resultant RAIDS with SATA, are at least two times faster. In my experience, working with a 500 MB image, drive delays using a two drive RAID, results in no more than a three second wait. Processing such images takes longer than that, so the wait is hardly a backbreaker.
I know of institutions that work with 2 GB images all of the time with 4 GB of RAM, and they aren't put off by it, though they would like to have more.
So, yes, multiple cpu's are more important than more RAM?at this time.
It's possible for a 32 bit program to handle larger images. The GIMP has done this for several years by transparently breaking up the working data into "tiles". When I remembered this, I realized that there is not a sufficient reason to port it to 64 bit yet.
Just 64-bit memory addressing would be a nice start - i.e., it's not necessary for the whole program to be based on 64-bit.
I don't disagree with your arguements, I just think Adobe is a monopoly that is milking it's user base. I'm sure it's a challenge to move long existing code to "universal/xcode" and 64-bit. If there was still a challenger (e.g., Macromedia), Adobe would be more motivated to do it sooner than later. The example of Quark brought up before is a good one.
No Kidding. I would like to see the breakdown of Mac vs. Windows *legal* CS owners.
There is a lot of anti-Adobe sentiment because Adobe has been playing games with their Mac user-base. Ever since they came out with that web page years ago saying that PC's were better tools than Mac's.... they've been steadily jabbing Mac users ever since. I don't know if Adobe has a burr up it's a$$ from iPhoto or FCP or what, but their attitude towards Apple and the Mac in general is typical of a company that wants to monopolize an industry for the sole purpose of having everyone by the short-and-curly's. They own the design software industry and they're showing it. Apple's MacPro sales have been affected by Adobe's lack of native support and I'd venture a bet that Chizen is clasping his hands and laughing maniacally. Bastard.
Despite the fact that we Mac users think the world revolves about us, it doesn't. Even if three quarters of the Windows installs are illegal, it still seems to result in almost three times as many legal installs.
few people here seem to understand that Adobe is not an arm of Apple.
In fact, it has been considered that Apple still exists BECAUSE of Adobe. The opposite is certainly not true!
Those running Adobe have to consider their own company. Apple, to them is, and should be, just another platform they write to. As Apple screwed up over the years, and saw its marketshare dwindle, Adobe did, what for them was the right thing, and concentrated on the larger part of their business, Windows.
There are complications here, to be sure. But Adobe can't favor the Mac when it consisted of just 27% of their business. If Apple did that, people here would scream.
Now that MS is mounting an attack on Adobe, Adobe is moving back a bit to Apple. But Apple is attacking slightly as well.
Adobe must take care of their own.
I hope that increasing marketshare for Apple results in more pros coming back to the platform. That will make things easier.
Despite the fact that we Mac users think the world revolves about us, it doesn't. Even if three quarters of the Windows installs are illegal, it still seems to result in almost three times as many legal installs.
few people here seem to understand that Adobe is not an arm of Apple.
In fact, it has been considered that Apple still exists BECAUSE of Adobe. The opposite is certainly not true!
Those running Adobe have to consider their own company. Apple, to them is, and should be, just another platform they write to. As Apple screwed up over the years, and saw its marketshare dwindle, Adobe did, what for them was the right thing, and concentrated on the larger part of their business, Windows.
There are complications here, to be sure. But Adobe can't favor the Mac when it consisted of just 27% of their business. If Apple did that, people here would scream.
Now that MS is mounting an attack on Adobe, Adobe is moving back a bit to Apple. But Apple is attacking slightly as well.
Adobe must take care of their own.
I hope that increasing marketshare for Apple results in more pros coming back to the platform. That will make things easier.
Just 27%...Adobe should abandon the Mac...I mean c'mon, just 27%.
The opposite, melgross, could certainly be true. Had Adobe abandoned the Mac a few years ago, it'd be in deep shit. In fact, at 27%, they'd still be in deep shit if they decided to drop Mac support today.
27% may not look like much to someone that doesn't know much about businesses but lemme tell you...abandoning 27% of your bread and butter customers can effectively kill a business or stop it dead in its tracks...Adobe certainly would certainly think twice about pouring R&D money into Lightroom and Soundbooth if 27% of your userbase was instantly wiped out.
Anyways...I don't care if Adobe abandons the Mac. It might hurt Apple in the short run but Apple would strike back in no time with a Pro app that does almost everything PS does and retain compatibility with PS. A lot might just migrate to Windows but if these guys have an Intel-based Mac, Apple can easily lure them back with its own app.
So, they'll rush out and buy CS3, but they'll leave Leopard on the shelf? OK, we get it. The smarties are telling us why 64-bit doesn't matter... yet. I get that. The point, though, is that if they were really pushing the envelope and not just cashing in on incremental updates, they would be there IN ADVANCE. It will matter.
Plus, "Pros and shops" don't act as a group. Some are early adopters and have the Mac Pros that can handle it, some are still using clusters of G5s that they invested so much in a while ago.
P.S. What Apple does with FREE programs is irrelevant.
The way the industry operates is that they never change anything in the middle of a campaign. when we had an ad campaign going, which could take months of work for us, we would stick to whatever we were using.
Just getting a new font in would send chills down our spines!
A new OS is the LAST thing anyone would want to move to. It takes at least 6 months before a new OS is stable enough to trust in a production setting. And yes, that means the Mac OS as well!
Programs need time to get their act together with the new OS.
It's far easier to try out a new CS, on a seperate machine, than a new OS AND a new CS version. Multiple unexpected problems coming from where? No one will know for months!
We've had far more printing problems fron Apple's OS upgrades, and sometimes from their updates, than from a new version of CS.
When Apple went to 10.3, our Fuji Pictography machines stopped working, and the OS would crash. Our scanners also stopped working, as well as other machines.
It turned out that Apple removed critical software from the OS in order to depreciate SCSI in favor of Firewire.
A stupid move indeed!
This caused many problems for the industry, and caused very few to move to 10.3. We had to reinstall 10.2 on several machines.
It also caused a move to Windows. MS almost never depreciated anything of this nature.
So don't talk about pro shops unless you know what you are talking about.
Perhaps, some of you should take a few moments and read Byer's official blog. It seems the comments and his responses are more educated and informative that what is appearing here.
Why should they bother when their feelings make them feel all warm and fuzzy?
Hmm. If it was that trivial to implement 64bit functionality into Photoshop then surely they would have done so already?
I know there is a plug-in which passes certain calculations through Altivec and there is an equivalent MMX/SSE plug-in for Windows. When the G5 came out it came with an updated version of the Altivec plug-in which was supposed to "take advantage of the power of the G5" of some such nonsense, but as far as I know all it did was add support for IBM's version of Altivec (VMX, or whatever it's called).
However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM.
The difficulty for Adobe is to update the way the core image, layer and blending code works. Not to mention their entire virtual memory management system. 64bit systems have been out for a while now, and I just find it irritating that despite the constant request from customers for the last 4 years to allow Photoshop to access more memory Adobe is brushing them off and giving us "new" features in CS3 which are either ripped from Photoshop Elements or Illustrator CS.
There is a difference between fairly easy and trivial.
Adobe is undergoing the biggest move they have ever made with CS3.
Incorporating Macromedia's programs, features, and at times, code, into their own major programs is not something they can do in a matter of a couple of months.
Moving to X Code is even more harrowing, despite Apple's public assurances. Look at how long it took from the quick demo of Mathamaticia to the ACTUAL program release.
Adobe does what is proper. 64 bit coding is something few need now. It's also something they can start moving to before CS4. Windows 64 has only been out in a meaningful way for about a year. Appl'es OS has not been 64 bit in any meaningful way.
Adjustment layers and Histry use very little RAM. A history uses not much more than a few KB's of RAM apiece. Adjustment layers use maybe a couple hundred KB's each.
The only time you really need more is when duplicating am image layer, or doing comps.
But, unless you are using multiple hundred MB size images, it isn't a problem for a machine with 4 GB's of RAM to handle.
I've rarely had memory swapping, even with very large images. Read my response to Sunil, above.
AS others have indicated it is difficult to give value to the logic of his argument when he is talking about an operating system that will be replaced by something that totally devalues his argument by the time his product comes out. Makes him sound like an Estate Agent. Worse still one can't but feel that om past experience and prsent promises a 64 bit version of photoship won't be around for a further year or so after CS3 launch when the equipment and software to take advantage of it will be quite mature to say the least. Fine if you want to be a follower but if Adobe don't want to lead the moves to superior computing at least be honest about it and accept that this is about money over innovation. Sounds very Microsoft to me which probably shows how they vewiw the world of competition and develop accordingly. A shame but a serious opportunity for others to compete with what Adobe think are dominant products.
You are also making the mistake by thinking that the date an OS comes out is the date when things change.
It isn't.
It will take months before even a small percentage of users will be on the new OS.
What makes you think that Adobe can feel confident that they know enough about the OS now, before it is out?
The likelihood that they have all the info they need is about zero!
It won't be until after the OS is out that they will be able to properly evaluate it. Complex programs need to delve more deeply into the OS in more places than a simpler program.
We don't know how much code is actually finished in 10.5. From what we see written about it, it still seems to have a way to go. No responsible company is going to try to rewrite all of a program at once, based on incomplete knowledge of a new OS.
Remember that the program also to continue to work on 10.4, and even earlier, though with diminished features and performance.
Comments
?Byer in his official blog pointed out that many of the perceived benefits of 64-bit computing simply won't manifest themselves with current generations of hardware and software.?
Perhaps, some of you should take a few moments and read Byer's official blog. It seems the comments and his responses are more educated and informative that what is appearing here.
I don't see as much of a problem as you do.
The program doesn't have to be 64 bit. That's a misconception. It does need to address more than 32 bits of memory. That isn't too difficult to impliment. They can do that later. The chipsets themselves don't address 64 bits for RAM anyway.
Most filters are already being handled by either Altivec on the PPC, or SSE on Intel. Both address 128 bits.
We don't need 64 bit menus and such.
Hmm. If it was that trivial to implement 64bit functionality into Photoshop then surely they would have done so already?
I know there is a plug-in which passes certain calculations through Altivec and there is an equivalent MMX/SSE plug-in for Windows. When the G5 came out it came with an updated version of the Altivec plug-in which was supposed to "take advantage of the power of the G5" of some such nonsense, but as far as I know all it did was add support for IBM's version of Altivec (VMX, or whatever it's called).
However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM.
The difficulty for Adobe is to update the way the core image, layer and blending code works. Not to mention their entire virtual memory management system. 64bit systems have been out for a while now, and I just find it irritating that despite the constant request from customers for the last 4 years to allow Photoshop to access more memory Adobe is brushing them off and giving us "new" features in CS3 which are either ripped from Photoshop Elements or Illustrator CS.
...However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM....
I agree. But how many pros can afford say 8gb-10gb Mac Pro systems? Intel Core2 [Conroe] tops out at 4gb. It would be a fairly elite group that has the RAM to go past that 3gb limit.
However, as I mentioned in my projections of RAM spec's becoming "standard" as the years go by, in 2009 probably is when we're looking at dumping the whole project into 10gbs of your 16+gb machine. CS4....
I fully expect Apple to come out with another app that chips away at Adobe's capabilities. They did it Final Cut, iPages, iWeb, Apeture, and Preview/PDF. So a more advanced image editing app than iPhoto and Apeture would seem natural (especially w/ CoreImage).
While that's all well and fine, I agree w/ an earlier post that Adobe's lineup is more universally accepted if not absolutely required for each field. Instead of developing a wannabe, Apple is better off taking over Adobe and controlling the farm.
The Acrobat line purchase alone would scare the heck out of Microsoft!
/
Let Apple just buy Adobe and squash this whole matter! Why shouldn't they?
Because competition is good for us?
wouldn't Apple benefit greatly is using their massive war chest to buy out Chizen's little monopoly?
Apple might benefit. Whether the users would is a wholly different matter.
I expect the real reason they don't want to add 64bit support is that it is going to be a major rewrite of the code and it'll be a nightmare to make it work on both Windows and Mac OS X.
Cross-platform compatibility and grotesquely old legacy code are the main things which are stopping development of Photoshop these days.
I agree with these comments. I'm sure Adobe's plate is full just transitioning to Mac Intel. To undertake a major rewrite of the very foundations of Photoshop would be too much to manage at the same time. And then there's that cross platform concern--about which I care nada, but Adobe does.
Chris
So, they'll rush out and buy CS3, but they'll leave Leopard on the shelf? OK, we get it. The smarties are telling us why 64-bit doesn't matter... yet. I get that. The point, though, is that if they were really pushing the envelope and not just cashing in on incremental updates, they would be there IN ADVANCE. It will matter.
Plus, "Pros and shops" don't act as a group. Some are early adopters and have the Mac Pros that can handle it, some are still using clusters of G5s that they invested so much in a while ago.
My understanding and experience is that people that do paying creative work with their computers are usually at least slightly conservative with their software, because of their past experiences of rushing into new software. They might buy the software but that doesn't mean that they will use it right away. For me, the released version of Tiger wasn't all that great until after a few updates, and I don't expect Leopard to be either. The pros will probably buy CS3 but will all work switch to it right away? Probably not except for the daring. The rest might experiment with it from time to time or try it for less time-critical projects to get a feel for its stability and acclimate to the changes.
P.S. What Apple does with FREE programs is irrelevant.
What are you refering to here?
JeffDM - "The filters wouldn't benefit from 64 bit code? I suppose that's all handled by vector ops now. But I thought people were begging Adobe to have something to handle images larger than 2GB for many years." You are right...
It's possible for a 32 bit program to handle larger images. The GIMP has done this for several years by transparently breaking up the working data into "tiles". When I remembered this, I realized that there is not a sufficient reason to port it to 64 bit yet.
Hi Melgross, Happy New Year 8) ... Actually I forgot my post several days ago somewhere else about 8 cores and Photoshop.
http://creativebits.org/8_core_mac_pro
Happy holidays to you over there across the world!
This (multiple cpu's), is actually far more important these days.
Adobe has always said that PS worked with multiple cpu's.
And they were right.
But, they didn't imagine that the word multiple would mean more than two!
So now they are scrambling to increase the number of threads. Not easy, but not impossible for the work the program is called to do.
Properly handled, it will work well with 3 to 4 GB's of RAM.
Someone here said that the program called for 7GB of RAM when rotating an image or some such operation. I can't imagine why it would. I've not been able to verify that on a smaller scale.
PS has always, according to Adobe, needed at least 3 times as much RAM as the largest image being worked on at the time for 100% efficiency, that is, all of the work being done in RAM, and as much as 5 times under extreme circumstances. The efficiency report, always available from the bottom of the current window, will let one know.
Working with many cores actually allows less memory to be needed at any one time, as the image is broken into smaller boxes. So there are other beneficial effects from going multiple core.
Of course, all this is mitigated by what plug-ins one has running at the time, and whether or not multiple images are open for use in comps.
But, rarely are people working on multiple 300 MB images at one time as we use to do in my place.
Even when we worked on 500 MB+ images for large billboards, the question of HD speeds rarely entered into the discussion. A fast RAID took care of most of that, and the delays were hardly ever more than a few seconds.
Today, drives, and the resultant RAIDS with SATA, are at least two times faster. In my experience, working with a 500 MB image, drive delays using a two drive RAID, results in no more than a three second wait. Processing such images takes longer than that, so the wait is hardly a backbreaker.
I know of institutions that work with 2 GB images all of the time with 4 GB of RAM, and they aren't put off by it, though they would like to have more.
So, yes, multiple cpu's are more important than more RAM?at this time.
It's possible for a 32 bit program to handle larger images. The GIMP has done this for several years by transparently breaking up the working data into "tiles". When I remembered this, I realized that there is not a sufficient reason to port it to 64 bit yet.
Just 64-bit memory addressing would be a nice start - i.e., it's not necessary for the whole program to be based on 64-bit.
I don't disagree with your arguements, I just think Adobe is a monopoly that is milking it's user base. I'm sure it's a challenge to move long existing code to "universal/xcode" and 64-bit. If there was still a challenger (e.g., Macromedia), Adobe would be more motivated to do it sooner than later. The example of Quark brought up before is a good one.
No Kidding. I would like to see the breakdown of Mac vs. Windows *legal* CS owners.
There is a lot of anti-Adobe sentiment because Adobe has been playing games with their Mac user-base. Ever since they came out with that web page years ago saying that PC's were better tools than Mac's.... they've been steadily jabbing Mac users ever since. I don't know if Adobe has a burr up it's a$$ from iPhoto or FCP or what, but their attitude towards Apple and the Mac in general is typical of a company that wants to monopolize an industry for the sole purpose of having everyone by the short-and-curly's. They own the design software industry and they're showing it. Apple's MacPro sales have been affected by Adobe's lack of native support and I'd venture a bet that Chizen is clasping his hands and laughing maniacally. Bastard.
Despite the fact that we Mac users think the world revolves about us, it doesn't. Even if three quarters of the Windows installs are illegal, it still seems to result in almost three times as many legal installs.
few people here seem to understand that Adobe is not an arm of Apple.
In fact, it has been considered that Apple still exists BECAUSE of Adobe. The opposite is certainly not true!
Those running Adobe have to consider their own company. Apple, to them is, and should be, just another platform they write to. As Apple screwed up over the years, and saw its marketshare dwindle, Adobe did, what for them was the right thing, and concentrated on the larger part of their business, Windows.
There are complications here, to be sure. But Adobe can't favor the Mac when it consisted of just 27% of their business. If Apple did that, people here would scream.
Now that MS is mounting an attack on Adobe, Adobe is moving back a bit to Apple. But Apple is attacking slightly as well.
Adobe must take care of their own.
I hope that increasing marketshare for Apple results in more pros coming back to the platform. That will make things easier.
Despite the fact that we Mac users think the world revolves about us, it doesn't. Even if three quarters of the Windows installs are illegal, it still seems to result in almost three times as many legal installs.
few people here seem to understand that Adobe is not an arm of Apple.
In fact, it has been considered that Apple still exists BECAUSE of Adobe. The opposite is certainly not true!
Those running Adobe have to consider their own company. Apple, to them is, and should be, just another platform they write to. As Apple screwed up over the years, and saw its marketshare dwindle, Adobe did, what for them was the right thing, and concentrated on the larger part of their business, Windows.
There are complications here, to be sure. But Adobe can't favor the Mac when it consisted of just 27% of their business. If Apple did that, people here would scream.
Now that MS is mounting an attack on Adobe, Adobe is moving back a bit to Apple. But Apple is attacking slightly as well.
Adobe must take care of their own.
I hope that increasing marketshare for Apple results in more pros coming back to the platform. That will make things easier.
Just 27%...Adobe should abandon the Mac...I mean c'mon, just 27%.
The opposite, melgross, could certainly be true. Had Adobe abandoned the Mac a few years ago, it'd be in deep shit. In fact, at 27%, they'd still be in deep shit if they decided to drop Mac support today.
27% may not look like much to someone that doesn't know much about businesses but lemme tell you...abandoning 27% of your bread and butter customers can effectively kill a business or stop it dead in its tracks...Adobe certainly would certainly think twice about pouring R&D money into Lightroom and Soundbooth if 27% of your userbase was instantly wiped out.
Anyways...I don't care if Adobe abandons the Mac. It might hurt Apple in the short run but Apple would strike back in no time with a Pro app that does almost everything PS does and retain compatibility with PS. A lot might just migrate to Windows but if these guys have an Intel-based Mac, Apple can easily lure them back with its own app.
Melgross:
So, they'll rush out and buy CS3, but they'll leave Leopard on the shelf? OK, we get it. The smarties are telling us why 64-bit doesn't matter... yet. I get that. The point, though, is that if they were really pushing the envelope and not just cashing in on incremental updates, they would be there IN ADVANCE. It will matter.
Plus, "Pros and shops" don't act as a group. Some are early adopters and have the Mac Pros that can handle it, some are still using clusters of G5s that they invested so much in a while ago.
P.S. What Apple does with FREE programs is irrelevant.
The way the industry operates is that they never change anything in the middle of a campaign. when we had an ad campaign going, which could take months of work for us, we would stick to whatever we were using.
Just getting a new font in would send chills down our spines!
A new OS is the LAST thing anyone would want to move to. It takes at least 6 months before a new OS is stable enough to trust in a production setting. And yes, that means the Mac OS as well!
Programs need time to get their act together with the new OS.
It's far easier to try out a new CS, on a seperate machine, than a new OS AND a new CS version. Multiple unexpected problems coming from where? No one will know for months!
We've had far more printing problems fron Apple's OS upgrades, and sometimes from their updates, than from a new version of CS.
When Apple went to 10.3, our Fuji Pictography machines stopped working, and the OS would crash. Our scanners also stopped working, as well as other machines.
It turned out that Apple removed critical software from the OS in order to depreciate SCSI in favor of Firewire.
A stupid move indeed!
This caused many problems for the industry, and caused very few to move to 10.3. We had to reinstall 10.2 on several machines.
It also caused a move to Windows. MS almost never depreciated anything of this nature.
So don't talk about pro shops unless you know what you are talking about.
Perhaps, some of you should take a few moments and read Byer's official blog. It seems the comments and his responses are more educated and informative that what is appearing here.
Why should they bother when their feelings make them feel all warm and fuzzy?
Hmm. If it was that trivial to implement 64bit functionality into Photoshop then surely they would have done so already?
I know there is a plug-in which passes certain calculations through Altivec and there is an equivalent MMX/SSE plug-in for Windows. When the G5 came out it came with an updated version of the Altivec plug-in which was supposed to "take advantage of the power of the G5" of some such nonsense, but as far as I know all it did was add support for IBM's version of Altivec (VMX, or whatever it's called).
However, that's only good for filters. When you're working on a couple of layered RAW images 500MB each with a few adjustment layers on top and 40 levels of history RAM usage is going to go up. Altivec & SSE are good for processing chunks of data at a time - ideal for filters, but if the program itself is swapping chunks of memory from RAM to the scratch disk every time you make a change on an open project that's going to limit performance. If Photoshop could instead store all of that in RAM rather than on the drive I'm sure you'd see a performance increase. It's not just about how much memory the processor can access at a time. As you say, we have 128 bit processing going on with Altivec/SEE. What I would like to see is reduced virtual memory swapping, and the only way I see that happening is if Photoshop is able to place the entire project you are working on in RAM.
The difficulty for Adobe is to update the way the core image, layer and blending code works. Not to mention their entire virtual memory management system. 64bit systems have been out for a while now, and I just find it irritating that despite the constant request from customers for the last 4 years to allow Photoshop to access more memory Adobe is brushing them off and giving us "new" features in CS3 which are either ripped from Photoshop Elements or Illustrator CS.
There is a difference between fairly easy and trivial.
Adobe is undergoing the biggest move they have ever made with CS3.
Incorporating Macromedia's programs, features, and at times, code, into their own major programs is not something they can do in a matter of a couple of months.
Moving to X Code is even more harrowing, despite Apple's public assurances. Look at how long it took from the quick demo of Mathamaticia to the ACTUAL program release.
Adobe does what is proper. 64 bit coding is something few need now. It's also something they can start moving to before CS4. Windows 64 has only been out in a meaningful way for about a year. Appl'es OS has not been 64 bit in any meaningful way.
Adjustment layers and Histry use very little RAM. A history uses not much more than a few KB's of RAM apiece. Adjustment layers use maybe a couple hundred KB's each.
The only time you really need more is when duplicating am image layer, or doing comps.
But, unless you are using multiple hundred MB size images, it isn't a problem for a machine with 4 GB's of RAM to handle.
I've rarely had memory swapping, even with very large images. Read my response to Sunil, above.
AS others have indicated it is difficult to give value to the logic of his argument when he is talking about an operating system that will be replaced by something that totally devalues his argument by the time his product comes out. Makes him sound like an Estate Agent. Worse still one can't but feel that om past experience and prsent promises a 64 bit version of photoship won't be around for a further year or so after CS3 launch when the equipment and software to take advantage of it will be quite mature to say the least. Fine if you want to be a follower but if Adobe don't want to lead the moves to superior computing at least be honest about it and accept that this is about money over innovation. Sounds very Microsoft to me which probably shows how they vewiw the world of competition and develop accordingly. A shame but a serious opportunity for others to compete with what Adobe think are dominant products.
You are also making the mistake by thinking that the date an OS comes out is the date when things change.
It isn't.
It will take months before even a small percentage of users will be on the new OS.
What makes you think that Adobe can feel confident that they know enough about the OS now, before it is out?
The likelihood that they have all the info they need is about zero!
It won't be until after the OS is out that they will be able to properly evaluate it. Complex programs need to delve more deeply into the OS in more places than a simpler program.
We don't know how much code is actually finished in 10.5. From what we see written about it, it still seems to have a way to go. No responsible company is going to try to rewrite all of a program at once, based on incomplete knowledge of a new OS.
Remember that the program also to continue to work on 10.4, and even earlier, though with diminished features and performance.