Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mac OS X Lion: Drops PowerPC Emulation, Adds QuickTime Pro Features, Much More
- Thread starter MacRumors
- Start date
- Sort by reaction score
Don't you mean
What make/model is the computer and how old is it? If your computer vendor can't even be bothered to provide an OS for it then it is a clear message to me that purchasing off that said company in the future isn't a smart idea.
How is it Apple's or Microsoft's responsibility to provide support for a device that they themselves never designed, built or sold? Maybe the 'rage' should be directed at Canon for failing to provide support for their devices than blaming it on Apple, just as you should blame the hardware vendors who create crappy drivers instead of blaming Microsoft because your scanner doesn't work.
Then doesn't that sound like a positive sign that maybe the hysterics are just that, hysterics?
Mate, I had to shell out upgrading from CS4 to CS5, I could have come on here wasting time venting my fury at Adobe and Apple but instead I sucked in my bottom lip, built a bridge and got over it.
As explained already it adds additional maintenance costs with little in the way of return. Screaming and yelling the same tired and pathetic argument over and over again won't some how change the underlying circumstances that led Apple to the decision it made.
The point still stands - you're adament that the whole world ceases all progress forward because YOU don't want to move - well sunshine, tough toe nails.
******** it 'quickly drops support' - come on, 4 f-cking years, 4 whole years you had the chance to upgrade! four years! four miserably long and pathetic years! You really are testing my patience.
Last edited by a moderator:
Somewhat lost in the ruckus is this little quirk of naming:
OS X 10.7 will be the first release that isn't compatible with actual OS X binaries. Specs that say "Requires OS X 10.4 or higher" are going to confuse a lot of people.
Perhaps it should just be called Mac iOS 5. Runs all iOS Apps (using Apple's magical Griselda technology) and Mac App Store apps.
Just jail-break it, and you can run your OS X 10.6 programs too.
You know it is coming.
OS X 10.7 will be the first release that isn't compatible with actual OS X binaries. Specs that say "Requires OS X 10.4 or higher" are going to confuse a lot of people.
Perhaps it should just be called Mac iOS 5. Runs all iOS Apps (using Apple's magical Griselda technology) and Mac App Store apps.
Just jail-break it, and you can run your OS X 10.6 programs too.
You know it is coming.
I am going to assume supporting x86 on an x86 system is pretty easy. It might add 1-2% over head to a program on memory calls. That being that the OS adds a extra memory offset for everything from a 32 bit app. Worse case it is might be an extra clock cycle when stuff needs to be access from ram but some how that is iffy since offsets are pretty much used for all ram in tracking where something is stored in memory.
1-2% extra overhead to any given program is not very much overhead. It might even be better to eat that extra over head than dealing with the larger size of a 64 bit program. Just have to look at the history of when we went from 16 bit to 32 bit. 16 bit was faster up until they ran out of memory and at that point 32 bit started taking over and 32 bit had more memory bandwidth it could use
x86 on x86 is trivial - that's native. Same with x64 on x64 - again, that's native.
The issue is x86 on x64 (that is, 32-bit binaries on a 64-bit OS). This involves more - such as switching the processor between 32-bit mode and 64-bit mode when making kernel calls, "thunking" kernel APIs to remap 32-bit APIs to 64-bit APIs, having a "parallel universe" of 32-bit shared libraries for 32-bit apps, ....
Once you've done the work though, it just keeps on ticking. Since the PowerPC architecture is dead (in regards to Apple OSX), freeze the APIs. Don't support any new APIs for PPC (support PPC development, though, so that bug fixes for PPC apps are possible).
Many people/companies have random bits of legacy PPC code in their workflows. With Apple's tens of billions in the bank, they should keep the promise that they made when announcing Rosetta along with the Intel switch. It's a noise expense.
The device works fine in 10.6.
The issue is that Apple is removing a much-ballyhooed feature in 10.7, and that causes the device to fail.
One can understand progress - that significant new features may require breaking some legacy code. In particular, moving to a new architecture (like moving from x86 to x64) can be expected to break low level code with architectural dependencies.
Just what does the user gain when Apple removes Rosetta?
So you agree that Apple should drop support for all x86 programs, and only support x64 binaries - since it's been 4 years since Apple has sold an x86-only system.
I believe porting x86 to x86-64 is a lot easier than going x86-64 to PPC.
Reason being is instruction calls for x86 and x64 are pretty much the same in assembly language. Same register set up and what not. Problem with PPC is is RISC and not CISC like Intel chips are.
That means very different instruction calls and as such a lot more over head for running rosetta.
I still think it was pretty stupid for Apple to drop it because it is not like they did not have all the work done for it. They are showing a little greed here is all I am seeing. The legacy application and I am going to guess fair amount of custom code is going to get screwed over. It is just yet another reason why Apple is not in the Enterprise world. They provide no road map and change things on a whim.
Thanks... I'll look into it-- it looks a bit expensive for a scanner that originally cost $100 or so, but it may be a better option than adding perfectly fine equipment to a landfill (especially if there are sales)
These seem like huge inferences and assumptions to me, without knowing me or others well. I'm not name-calling or belittling you or anyone else here.
Enjoy Lion.
Nobody (I mean a vanishingly small number of people) codes in assembly anymore. Whatever language your using has compilers to generate the assembly code from your high level code.
And nobody (or vanishingly small) is porting from x64 to PPC. In the AIX world, there's a lot of POWER -> x64 porting going on.
The issue here is that many people have legacy PPC OSX applications. For whatever reason (app has been end-of-lifed, company went out of business, don't want to pay the upgrade fee for the x86/x64 version,...) they have been happy with Rosetta emulation of PPC on Intel for the apps that they've been using for years.
Now it seems that with little justification Apple is removing a feature from Apple OSX (that is, downgrading). That will cause these Apple customers pain.
An astute observation...
Apple switched to Intel almost 6 years ago( announced at the WWDC June 2005 ), so the final vanquishing of PowerPC support is expected IMO. Moreover, if someone has software that is not universal, you need to contact your vendor and find out why 5+ years wasn't enough time for them to produce a universal/intel binary. Clearly, Apple does not want to compile their frameworks for PPC going forward and that will obviously stop Rosetta from running. Folks who have old apps that require Rosetta should just keep a mac running OS 10.6 or lower. No problem. Apple needs to move on and the move makes sense and they've provided more than adequate notice to developers.
Sorry not talking about people coding it so to speak. I mean more for a compiler to translate it. By that I mean the amount of overhead require for the computer handle translating PPC code to x64.
Now getting x86 code to work on an x86-64 computer is easy since well AMD figured out how to get both on a chip and let the CPU handle it so to speak. Requires a lot less overhead to do it.
I can tell you from my classes Assembly language sucks to code in but handy to have a least some type of understanding of it.
I suggest you look at the cost of doing work on custom code. If you have custom code that you paid tons of money for you sure has hell would not want to pay it again. Custom coding is very expensive. Give you an idea custom code starts at around $100 per man hour of work and only goes up from there.
Indeed. Most x86 code should run without modification on AMD64. Legacy mode is entirely compatible. In long mode we removed a few things that nobody uses.
In other words, we use the same opcodes and operand formats, and simply added new stuff for 64-bit support. Code doesn't need to be recompiled at all unless you want it to be 64-bit, and even then it's fairly trivial.
So why you said that you "will be pushed towards having to jump to a PC" if you intend to stay on your current Mac?
So what you're saying is that people should just pony up hundreds of dollars to upgrade their perfectly working software so Apple doesn't have to bother to support their own operating system? The issue isn't that it's not available in some cases (like Photoshop), but rather that what will probably be a $30 OS upgrade (given its lack of 'major' features) can cost people hundreds of dollars in upgrades that worked fine on the same computer with the earlier OS incarnation. Given Lion dumps the "grid" for Spaces (I don't like that at all), adds no real value features (I don't give a crap about Launch Pad since I don't plan to use my Mac as iOS Mark II) and cripples some of my otherwise perfectly working software, I don't plan to "upgrade" to it at all. The problem is that some future software likely will require Lion to function, thus creating the paradigm of darned if you do and darned if you don't.
I have yet to hear one valid reason why Lion should not offer Rosetta support. I've seen nothing in the operating system that suggests it is somehow incapable of running their PPC emulator. I'm wondering if just manually copying it over would work, for that matter (or maybe a few tweaks/hacks). It's ridiculous to make software non-functional without a clear benefit to doing so. They could easily offer a "use at your own risk" banner if they don't want to worry about any support issues, but just dropping it is a deal breaker here. I simply won't install Lion on my MBP. Honestly, Snow Lion was of no benefit either as far as I can tell (XBench tests slower in almost everything compared to Leopard, which tests slower than Tiger in most areas. OSX is going backwards whereas up through Tiger, it got FASTER with every release).
Maybe you'd care to donate a couple hundred bucks so I can update to CS5?
Of course it isn't Apple's fault that the vendors aren't updating their software, and no one here is saying that. What we are saying, is that it's Apple's fault for not letting people use software they bought just a short while ago. Not all PPC software was made 5 years ago. Less than 6 months ago, I bought a program (and it cost several hundred dollars), and only now did I find that it is a PPC app. That isn't the only thing. It turns out that CS3, which I spent about a thousand dollars (AUD) on, just to find it has PPC code in certain places. I am not going to buy the new CS5 and spend more on it, just because Apple didn't bother supporting less than a year old software. And I know what lots of you are going to say...
No, he/she says that when it comes the time to buy a new computer, he would buy a Windows PC, not a Mac.
No, he/she says that when it comes the time to buy a new computer, he would buy a Windows PC, not a Mac.
Nobody forces you to upgrade, and SL won't magically stop working either.
So much whining.
If Rosetta is such a low-maintenance application, adjusting SheepShaver to run on Lion should also be a simple task. And copying the Rosetta binary over to Lion should essentially also just work.
And Rosetta certainly is not more dependent on frameworks and APIs than other apps like TextWrangler. Thus 'porting' Rosetta over to Lion will not require any considerations when modifying frameworks and APIs in Lion.
Is Apple making any money by not supplying Rosetta anymore? Are we buying new Macs that only run Lion (or buy Lion separately) if we have apps that do not run under Lion?
Certainly no. On the contrary, Apple is losing money by people holding back on upgrades (and by resentment building up). Their decision to not support Rosetta anymore thus bring them other financial benefits, lower development costs and a better OS (by not having to support legacy APIs etc.). But as you say, these must be negligible maintenance cost and the real reason is that Apple just wants to impose its worldview that PPC app are simply not chic anymore.
First of all, I said low maintenance, not no maintainence. That's relative to the size of the project naturally and OSX is a big one. You're kinda forgetting that software's unique in that once the coding is done there aren't very many physical resources used up in its distribution. Every sale that they wouldn't otherwise get is mostly money in the bank. Let's say there are only 10,000 people left to retain still using Power P.C. software, 1/2 of which would upgrade to Lion specifically, as you've said they're worth something. However your stipulation that Apple's both gaining and losing money isn't quite so true; while the resources shifting behind the scenes are rather complex, at the end of the day it either comes down to net gain or net loss.
You could afford to hire a dedicated team of ten guys working two full fortnights at $300 dollars an hour from 9-5, it'd cost 336,000 dollars to code the software but assuming the cost of the upgrade, is $130 dollars, as it usually has been in the past, you gain 650,000 dollars in sales, minus the cost of media, which assuming a cost of $10 per package brings us down 50,000. We've still got $214,000 left in the budget. Say shipping and handing costs us 5 dollars a package, that's $20,000 more so we've got $196,000 in profit from this, assuming everything works nicely with Lion's new tech which was of course a precondition of my prior post. If something in Lion makes PPC support infeasible, please, by all means drop it. That is after explaining why to your long term repeat customers of course. If you need to you need to but this idea of planned obsolescence in four-five year intervals simply bothers me as a customer of 'premium' products.
Secondly, Apple has lower level access to the resources necessary to maintain the O.S. something that might be relatively simple to the guys in Cupertino, might be increasingly complex to any benevolent hacker elsewhere who wants to pick up where Apple left off. If the only thing that's needed is a find/replace all of some odd string in the source code, who has access to the source code to actually go through with it? You've also got copyright issues to consider. I kinda doubt Apple would appreciate somebody distributing hacked versions of OSX.
Thirdly, uh people do similar work all the time, during hobbyist projects no less! Name a legacy video game console, any currently legacy video game console. The chances are that some way, somewhere, there's an emulator working for it on several home computer systems, many of which are free of cost. Aside from that, Apple themselves maintains a compatibility layer of similar nature to Rosetta in X11, albeit much of the work is already by the open source community. Albeit on the other hand, the only program I really run in X11 is The Gimp. Why don't they drop their implementation of X11, which allegedly has superior FoSS alternatives and dedicate those resources to Rosetta? There's also Wine and its commercialized cousin Crossover to consider.
As par why Apple might be doing this now, I'm not sure with the way the discussion's going. I'd guess it's probably licensing fees, coupled with a desire to push people to the App Store as evidented by the change to their download page. Apple stands to gain commissions from app store sales. Isn't there a focus on the App Store with Lion? I'm growing weary with this trend...
Somebody could surely make something that does convert the entire app based on similar principals though, couldn't they? I guess it's not the most transparent solution but if the install was self-contained in such a fashion, kinda like what Winebottler does?
Under janil's reasoning, you don't even have to know PowerPC processers ever existed to care, at least not until it's too late. Legacy peripherals and software still exist in the second hand market, retailer's backstock and old websites. If the technically inept see something listed as OSX compatible, they want to stick it in their Mac and see it just work. If it doesn't you might hear something to the effect of:
"My computer's like, uh broken. Can you come over here and fix it?"
"Sorry, I can't do anything about it. This program isn't compatible with your operating system.
""Yes it is! It says so right on the box!"
"Sorry, Apple blah blah blah."
"Huh?
"Optional repeat and addendum:
"I said blah blah blah."
"What? This is an outrage! I paid $____ dollars on this new _____ to get it working!
"While it might be an inevitable problem, that's not user friendly and should be avoided when possible. Besides for those with more savvy, the OSX software and peripheral library is already small enough as it is without diminishing it further. Here's to hoping Thunderbolt picks up soon to diminish this issue, at least on the hardware side of things.
Also with regards to no comment in particular, is legacy support really so resource intensive? I can understand how one might think that the transition from OS9 to OSX might've been based in dropping dead weight but we've got to remember, OSX has a "Rock Solid Unix Base" and Unix has been around for about 50+ years by my last count. I'm not sure but I think we may've inherited more legacy support than we dropped in the transition, particularly since I hear some of the common utilities are ancient. Besides that, many flavors of Linux like say Ubuntu, have entire PowerPC builds, at least to my understanding.
As par bloat, operating systems grow larger and larger all the time. I remember when my System Folder was only 90-100 megabytes big in System 7.5. The software's eventually going to grow with each new feature no matter what. To me, Rosetta's a feature like any other until it makes a mess of itself. Besides, if the only issue is bloat, everything extra could just be migitated to an installation disc as an optional install, so people who don't want to waste MBs or GBs wouldn't have to.
Last edited:
Nobody forces you to upgrade, and SL won't magically stop working either.
So much whining.
I wonder how many times people have to point out that many OSX developers (this isn't a big problem in the Windows world, but it's a HUGE problem with Apple operating systems) stop making software compatible with older operating systems relatively quickly (especially things like video compression and related apps like Handbrake, although even Firefox 4 won't support PPC Leopard any longer due to a lack of developer tools for newer technology from Apple and thus whether the OS 'works' or not is 100% MOOT) and thus your lack of upgrading will have a very real impact on your ability to use Snow Leopard, even if it doesn't "magically stop working". This has been pointed out numerous times in numerous threads, so much so that I wonder whether the people that insist on repeating that same tired line about not upgrading are purposely being obtuse.
Presumably( if not it would be pointless ), Apple is pulling the PowerPC code from their frameworks. Even if Rosetta were installed, it wouldn't run.
It seems your issue is with the developer of CS5 that didn't upgrade their app in the 5+ years since Apple announced the Intel change.
My clients( and I've been in this business for 35+ years ) all received the source code for their custom development and thus had the choice to recompile it for another platform. Again, if anyone has purchased a custom app that the developer failed to upgrade to Intel, they should contact the developer and/or read the contract with the developer to find out what the agreement was. Regardless, it still points to an issue between the end user and the developer and not an Apple issue. My clients received free upgrades to the Intel versions of their apps. It was/is the right thing to do. Quite frankly, I'm surprised Apple kept Rosetta around this long.
Exactly. So all of you with critical apps with PowerPC code:
(1) Keep Snow Leopard on at least one mac and run the apps there.
(2) While doing #1, contact the developer and ask them why 5+ years wasn't enough notice to provide an Intel version of their code. Regale them with your issues and insist on an Intel version.
(3) While you are doing #2 be prepared to pay for the upgrade because moving code from PowerPC to Intel is not automatic and takes work.
The real issue is Apple
The real issue is that Apple did not give notice that PPC support was a temporary bridge, and did not warn the developers years in advance when the EOL date would be.
You seem to imply that laziness is the cause, whereas I suspect that there's a big dollop of conservative "don't fix what ain't broke" engineering.
Why risk introducing new bugs by porting things that are working just fine? (At least, in the absence of any notice from Apple that support would be ending.)
The real issue is that Apple did not give notice that PPC support was a temporary bridge, and did not warn the developers years in advance when the EOL date would be.
You seem to imply that laziness is the cause, whereas I suspect that there's a big dollop of conservative "don't fix what ain't broke" engineering.
Why risk introducing new bugs by porting things that are working just fine? (At least, in the absence of any notice from Apple that support would be ending.)
Presumably, someone that comments on Photoshop might know something about it. The problem is not that Adobe hasn't updated to Intel code with CS4 (let alone CS5), but that it's not exactly free. Because Apple doesn't feel like supporting Rosetta, I can either keep Snow Leopard on my MBP or I can spend $200 to upgrade Photoshop (plus whatever it costs to upgrade to Lion). I'm not using Photoshop as much as I used to so it's a very real (and largely pointless cost) and yet Apple is deciding I should pay Adobe for an update I otherwise do not need. Compare this to Photoshop 6 I have for Windows and it still works fine three major operating systems later (i.e. it would be like if OS9 versions of Photoshop still worked in Lion...what a laugh that would be). The point is if you want your expensive applications to have a decent shelf life, don't buy them for the Mac.