Why Apple Must Continue G3 Support in Mac OS X 10.5 Leopard

We’ve been having quite a discussion about G3 support in Mac OS X 10.5 Leopard in the Low End Mac Mailbag, but I think this exchange merits a bit more attention. Kris Finkenbinder has some important things to say about the need for Apple to continue support of G3 Macs in the next version of the Mac OS.

Dan, first of all great site. I’ve been reading for a few years now and it’s been very helpful many times.

14" ibook G3I have to voice my protest over your general attitude toward whether or not Apple should continue supporting G3 Macs with Leopard. For one thing, you seem to find it easy to reference and then dismiss the fact that the last G3 iBooks are barely five years old now. If they aren’t supported by Leopard, that means they will have only a single Mac OS X upgrade version available before being forcibly, and unnecessarily, obsoleted. If I owned one of those machines, I wouldn’t find that acceptable. How can a switcher justify investing a huge chunk of cash in a Mac system when he can’t be assured that it will be able to run the current version of Mac OS X just five or six years down the line? That really kills the resale potential when something can actually be called obsolete with good cause due to a mere artificial software limitation.

Comparisons to Windows support on old hardware are disingenuous. Fairly old PCs can still run Vista with all the useless bells and whistles turned off, going back at least ten years. Will it run well? Probably not, but if someone has a particular use for a piece of software or feature that only runs on Vista and doesn’t require a powerful CPU, there is nothing stopping them from doing that. In contrast, Apple’s installer will absolutely refuse to install on “unsupported” hardware without being modified, even if the only real issue is that the unsupported hardware is a little slow. That’s just not good business practice. Making loyal users of older hardware feel abandoned is a really bad idea. Apple should support all Macs going back at least ten years, with warnings on older hardware, not complete blocks.

Mac OS X 10.4 TigerYou seem to have failed to notice that a large number of very useful pieces of third-party software are already incompatible with 10.2.8 and even 10.3.9, simply because the developers don’t know how or can’t be bothered to install the cross-development packages and check some boxes in Xcode. This started almost immediately after Tiger was released, and the same thing will probably happen as soon as Leopard is out, even though many of these incompatible applications aren’t doing anything that actually require 10.4 or 10.5. The only way to get around this lax developer attitude is to be able to install Leopard on all Macs, including G3s, to maintain maximum software compatibility.

The fact is that people are successfully running OS X all the way up to Tiger on much slower machines (at least as slow as 233 MHz, if not slower), and there is no real reason that Leopard shouldn’t be able to run just as well on those same machines.

You place too much emphasis on having a machine be able to run Leopard quickly with all the features turned on, thus your fairly arbitrary 500 MHz cutoff point. The fact is that people are successfully running OS X all the way up to Tiger on much slower machines (at least as slow as 233 MHz, if not slower), and there is no real reason that Leopard shouldn’t be able to run just as well on those same machines. People will find a way to disable any features that are particularly resource-intensive and/or unnecessary for daily use. Even here on my final-generation iBook G4 I have Dashboard disabled, because I just don’t need it taking up any resources. It is probable that at a bare minimum Leopard could be used on a G3 for general office type tasks and Web/email without anyone noticing any real performance problems. The deciding factor is mainly the availability of enough RAM, and even the most limited G3 can handle 512 MB. That may not be enough for what you like to do with your Macs, but for a lot of folks it’s just fine.

The hardware capabilities of G3 machines should be considered an entirely separate issue. I have a Power Mac 8500 that is perfectly capable of handling Panther and Tiger. With room for 1 GB of RAM and slots for installing modern hard drive controllers, gigabit ethernet, and FireWire 400/800 cards, these types of machines can be adequate not just for simple desktop use but they can be one of the best and cheapest available options for file servers. The CPU upgrade market seems to still be going strong, with speeds constantly increasing and even G3-based upgrades still being manufactured.

You dismiss upgrades to old Power Macs because they are expensive and a Mac mini is cheap. But a very capable G3 CPU upgrade can be had for as little as $180, without going into the $699+ ridiculous range that would justify junking the old Mac in exchange for a Mac mini. But due to the fact that old Power Macs can accept FireWire 800 and SATA/eSATA cards, they can actually be better choices for server use than a Mac mini, despite the expense!

Having support for Leopard and Leopard Server will make them even more useful. There is no real reason why a G3 Mac can’t be an excellent file server, Web server, wiki server, iCal server, software update server, Spotlight server, streaming QuickTime server, mail server, Time Machine network backup location, so on and so forth. Most server activities mainly involve pushing data through a network card with very little processing power necessary, and even in the few small offices with gigabit networking a fast hard drive will never be strained to keep up. With an SATA or ATA133 card even the largest hard drives can be supported, so disk space is not a problem either.

Furthermore, the ability to do piecemeal upgrades of a tower over the years is a lot easier for some people than forking out $600-800 all at once for a new Mac mini. Most PCI cards are down in the $40-80 range. I flat out don’t think you should be using the Mac mini as an excuse to help make it okay for Apple to abandon G3s. It’s a machine with limited usefulness to many people. For those who want to do video and such, sure, recommend the Mac mini. But to use it as a justifier for ending G3 support? No, I’d really rather you didn’t do that anymore. Even someone buying a used G3 right now and doing the connectivity upgrades can end up with a machine that is better at certain tasks than a Mac mini. Of course they would be better off with a G4 tower, but I’m just trying to make it clear that the Mac mini doesn’t relate well to the discussion of continued G3 support.

I think XPostFacto is also less relevant to this issue than you imply. If I understand things correctly, the only real problem with using OS X on pre-iMac G3 hardware was that the older boot ROMs weren’t designed to boot OS X and can’t have both OS X and OS 9 boot support enabled simultaneously. So all they had to do was build an app to tweak the boot ROM and use a slightly modified version of the open-source kernel. A G3/G4 upgrade card and everything runs all the way up to Tiger 10.4.9. No problem. Once the kernel boots there are no further software limitations, since Tiger was still made to run on a G3 processor.

…if Apple decides to actually stop compiling support for G3s into not just the operating system kernel but all the basic software that comes with it, solutions like XPostFacto won’t matter.

But if Apple decides to actually stop compiling support for G3s into not just the operating system kernel but all the basic software that comes with it, solutions like XPostFacto won’t matter. Even if the Leopard kernel will eventually be available as open source, and even if they can get the kernel to boot and skirt the installer limitations, a G3 system will wind up being almost useless because all the native Apple software will either be unstable or entirely nonfunctional. At least, any software that includes even the tiniest function call that requires a G4 or above processor. Admittedly I am not an expert in this area, but this is the conclusion I have come to based on a lot of reading. It appears that a total lack of G3 support would be a much more complicated barrier than the previous hurdles XPostFacto has dealt with. The support for G3s needs to come directly from Apple or it will probably be pointless to even try adding it on afterward.

Just my 2 cents, of course. You should know all this technical stuff as well as I do, since I picked up most of it from your site. Again, thanks so much for having one of the most useful Mac sites on the net. I am taking this time to write to you in depth on this issue because you are a widely heard voice in the Mac community. Because of that the things you say have more visibility than most, and could have a real impact on the course Apple finally chooses with Leopard. Please help us all by being a little more positive about the necessity for Apple to continue supporting all G3 Macs.

Sincerely,
Kris Finkenbinder, Owner
Red Bear Network Services
Sitka, AK 99835
Switcher and Mac promoter


 

Kris,

Please reread my comments. I’m not opposed to Apple supporting G3s in Leopard. In fact, I support it.

It’s reasonable to assume that Leopard will also leave a few older Macs behind, and we’re trying to determine a reasonable cutoff point.

Apple has been dropping support for some older Macs as OS X progresses. The OS X Beta didn’t require a G3 Mac, later versions required a G3 motherboard (no more support for upgrade cards), and Tiger is only supported on G3s with FireWire. It’s reasonable to assume that Leopard will also leave a few older Macs behind, and we’re trying to determine a reasonable cutoff point.

Let’s look at the oldest desktop and notebook Macs supported by each Mac OS X version:

I don’t think it’s unreasonable to believe that the 2007 release of Leopard will up the ante a bit. It would be consistent with past releases if Apple supported all hardware up to four years old and some hardware up to six years old.

There are different ways Apple could decide to make the cutoff. They could do it strictly on clock speed, the amount of video memory, a certain graphics processor, presence of FireWire ports (as with Tiger), a minimum screen size (probably ruling out the clamshell iBooks), etc.

In my opinion, AGP graphics and 1024 x 768 resolution is a reasonable minimum for Apple to support on OS X 10.5. That means including all slot-loading G3 iMacs and all dual USB G3 iBooks. The Blue & White Power Mac G3 and PCI graphics Power Mac G4 wouldn’t make the cut either with their PCI graphics, but the 1999 Sawtooth Power Mac G4 would, as would the 2001 PowerBook G4.

Some of these models have only 8 MB of video memory, and Quartz Extreme video acceleration requires 16 MB of VRAM, so there’s an outside chance Apple might make the cut there. That would unfortunately rule out the first generation PowerBook G4, early dual USB iBooks, and pre-2001 iMacs while supporting the 2001 PB G4/550 and 2002 iBook G3.

Thus Apple would be fully supporting all Macs up to five years old and some Macs up to eight years old if Leopard required AGP graphics and 16 MB of VRAM.

As for using older Macs as servers, you probably already realize that Mac OS X 10.3 Panther is probably the best for that job, as it has much lower system requirements and makes less demands on the CPU than Tiger does.

Yes, there will definitely be Leopard-only apps, as there are features in Leopard that programmers will want to take advantage of – just as they did with each previous version of OS X. It’s really their choice if they want to forgo support of earlier versions of OS X.

One key idea at Low End Mac is that no Mac is ever less capable than it was when you bought it, and future versions of the Mac OS will usually add new capabilities with some added overhead. For the most part, we think 10.3 is best for most G3 Macs and older G4s, while 10.4 is fine for the faster G3s (say 400 or 500 MHz and beyond) and most G4s.

Our advice to choose a Mac mini over a Power Mac G4 was specifically for someone involved in video work. The amount of raw processing power in a Core Duo CPU puts even a dual-processor G4 Power Mac to shame.

As always, the key is finding your balance point: How much CPU horsepower, how much memory, how big and fast a hard drive, how powerful a graphics processor, how much internal expandability, and what OS features do you need? That’s a personal decision, one which we try to help people make by offering information and advice.

But we have no input with regards to where Apple will make the cutoff for Leopard. We’re simply speculating based on historical trends. We may find out as early as June where Apple is drawing the line, and we can be pretty sure that Ryan Rempel will do all in his power to support Mac OS X 10.5 on as much older, unsupported hardware as possible with the next version of XPostFacto.

Dan


Kris replies:

Dan Knight wrote:

Kris,

Please reread my comments. I’m not opposed to Apple supporting G3s in Leopard. In fact, I support it.

I know you do support it – it’s just that you have so many qualifications going along with your support that your support seems very lukewarm as opposed to strident.

Apple has been dropping support for some older Macs as OS X progresses. The Beta didn’t require a G3 Mac, later versions required a G3 motherboard (no more support for upgrade cards), and Tiger is only supported on G3s with FireWire. It’s reasonable to assume that Leopard will also leave a few older Macs behind, and we’re trying to determine a reasonable cutoff point.

[snipped …]

I don’t think it’s unreasonable to believe that the 2007 release of Leopard will up the ante a bit. It would be consistent with past releases if Apple supported all hardware up to four years old and some hardware up to six years old.

Sure, perfectly reasonable that “official” support would leave some machines behind. Any reasonable person would agree. Nobody expects Apple to officially on paper support ancient machines. But that’s not what I was talking about. Tiger isn’t supported, but it works just fine on any Mac with a G3, FireWire absence notwithstanding.

There are different ways Apple could decide to make the cutoff. They could do it strictly on clock speed, the amount of video memory, a certain graphics processor, presence of FireWire ports (as with Tiger), a minimum screen size (probably ruling out the clamshell iBooks), etc.

In my opinion, AGP graphics and 1024 x 768 resolution is a reasonable minimum for Apple to support on OS X 10.5. That means including all slot-loading G3 iMacs and all dual USB G3 iBooks. The blue & white Power Mac G3 and PCI graphics Power Mac G4 wouldn’t make the cut either, but the 1999 Sawtooth Power Mac G4 would, as would the 2001 PowerBook G4.

What you seem to be talking about thus far is “official” support in the installer, whereas I was talking about the fact that they can’t drop G3 support without dropping support for all G3 processors. Since the code is the same for all generations of G3, AFAIK. If they compile in G3 support, it’s there, otherwise it just isn’t. Having such an arbitrary cutoff point so close to the end point of the G3 range makes very little sense for Apple. It would be a lot easier for them to simply leave all G3s behind. Promoting this cutoff point is thus an overall negative argument, I feel.

Clamshell iBooks and other 800 x 600 machines are already useless for certain applications like iLife, which require 1024 x 768. It’s annoying and arguably unnecessary, but lack of support from such display-oriented applications is a lot more acceptable than a hard screen size limit in the operating system would be. What about people who are using OS X on a machine that is operating as a server or in some other specialized capacity (e.g. a 640 x 480 display on a projector)? They can’t just go around restricting the whole operating system to a certain type of display or screen size. Therefore I submit that this is yet another arbitrary cutoff point that should have nothing to do with your reasoning.

Some of these models have only 8 MB of video memory, and Quartz Extreme video acceleration requires 16 MB of VRAM, so there’s an outside chance Apple might make the cut there. That would unfortunately rule out the first generation PowerBook G4, early dual USB iBooks, and pre-2001 iMacs while supporting the 2001 PB G4/550 and 2002 iBook G3.

It’s already been shown to be relatively simple to disable Quartz Extreme and the fancy effects that require it, thus making it extremely viable to use OS X on very old machines. This also applies to your additionally arbitrary cutoff of machines with AGP. If the user’s applications don’t require heavy graphics acceleration there is no reason that a PCI card won’t work just as well, just as it has up to this point. Those effects are just eye-candy like the Aero features in Vista, and shouldn’t have any bearing on the basic viability of the operating system on older hardware.

Thus Apple would be fully supporting all Macs up to five years old and some Macs up to eight years old is Leopard required AGP graphics and 16 MB of VRAM.

One of the reasons I can be so enthusiastic about promoting Apple as a computing alternative is the longevity of their machines, and not just hardware-wise.

As I’ve already mentioned, I am among the crowd that would find it unacceptable to have a 5-year-old computer be “officially” abandoned. One of the reasons I can be so enthusiastic about promoting Apple as a computing alternative is the longevity of their machines, and not just hardware-wise. The fact that I can still run Tiger on an officially supported original iMac that was manufactured before OS X was even released, is a tremendous positive effect. My confidence in pushing someone to make the usually sizable investment required to switch to the Mac platform would be seriously shaken if Apple took such a stand. Five years is nothing in a Mac lifetime; that’s one of the strongest selling points.

Without the capacity to run at least 10.3.3 (when the networking process finally started working smoothly), the usefulness of those machines would be almost nil, but since they can run the same OS version as every other Mac they can coexist gracefully with common software and networking abilities. They may not run Aperture, but they are just as good at Office 2004, Mail, iTunes, Firefox, and Safari as the next Mac (with sufficient RAM installed, and everything after the first iMac models supports a very capable 512 MB or 1 GB that is more than good enough for basic office use). Being able to run the exact same OS and software on every Mac not only vastly reduces administration headaches (ARD3 already requires 10.3.9 on clients), it lets the users move between systems without getting confused over interface idiosyncrasies. This is unbelievably useful for any Mac network from two computers to 2,000 computers. Family packs of OS X even make maintaining a common version an inexpensive and viable option for home users with multiple Macs. I think that was kind of the point of having a Family Pack.

As for using older Macs as servers, you probably already realize that Mac OS X 10.3 (Panther) is probably the best for that job, as it has much lower system requirements and makes less demands on the CPU than Tiger does.

As I pointed out before, most server processes focus on moving data around. Mail, iCal, wiki, contacts, and many other functions are all basically text file manipulation and require very little processor power unless you have quite a large number of users (hundreds and up). The main bottlenecks with servers are disk speed, network speed, and RAM, not the CPU. Obviously any organization with a large enough user base to strain a server will have the cash for getting a better machine. But there is no real reason that tens of thousands of small offices and even families can’t slap a 10-user copy of Leopard Server on any old Mac with an ethernet port and get all those new server features without needing a high-powered CPU. They aren’t just frivolous features either. A lot of offices could benefit from the improved file services, the latest Apache Web server, iCal/wiki servers, and improved iChat server functionality.

Yes, there will definitely be Leopard-only apps, as there are features in Leopard that programmers will want to take advantage of – just as they did with each previous version of OS X. It’s really their choice if they want to forgo support of earlier versions of OS X.

The only way to stay current if Safari is your chosen browser is to be able to run the current version of OS X.

I’m sure you’ve also noticed that even applications as important and useful as Safari are unnecessarily tied to your version of OS X. As the Web evolves, those frozen older versions of Safari are getting further behind. The only way to stay current if Safari is your chosen browser is to be able to run the current version of OS X. Since the ability to browse the Web efficiently and be compatible with new websites is one of the primary purposes of all computers, I would say that the inability to run the latest software for such a basic task is a very bad thing, and in the case of Safari that means you need the latest OS X. It’s just lucky that there are alternative Web browsers that continue to support previous versions of OS X with their latest builds, otherwise old OS X versions could end up being just as useless as OS 9 is now that there is only a single viable Web browser still being developed.

One key idea at Low End Mac is that no Mac should ever be less capable than when you bought it, and future versions of the Mac OS will usually add new capabilities with some added overhead.

Commendable, but the selling point for me is that new versions of Mac OS X have always added new capabilities, and most of the time also reduced overhead. Tiger is arguably the first version of OS X that has run a bit slower on certain old machines. But on machines with sufficient RAM it was once again faster than Panther, thus enhancing every Mac user’s investment and making it more likely that they will continue to purchase and promote Macs.

For the most part, we think 10.3 is best for most G3 Macs and older G4s, while 10.4 is fine for the faster G3s (say 400 or 500 MHz and beyond) and most G4s.

This is the attitude that I disagree with. What is best for the job depends entirely on the user. It seems you are always limiting the requirements for your definition to the idea of having an acceptably “snappy” desktop experience comparable to when the machine was new. If that is your main viewpoint, it might be good if you made that more clear. Many people are more than happy to put up with a few beach balls if it means they can stay current with basic productivity and web browsing software.

Our advice to choose a Mac mini over a Power Mac G4 was specifically for someone involved in video work. The amount of raw processing power in a Core Duo CPU puts even a dual-processor G4 Power Mac to shame.

As always, the key is finding your balance point: How much CPU horsepower, how much memory, how big and fast a hard drive, how powerful a graphics processor, how much internal expandability, and what OS features do you need? That’s a personal decision, one which we try to help people make by offering information and advice.

I agree, it is a personal decision. Unless Apple doesn’t compile in G3 support at all, which will take the decision out of our hands completely.

But we have no input with regards to where Apple will make the cutoff for Leopard. We’re simply speculating based on historical trends. We may find out as early as June where Apple is drawing the line, and we can be pretty sure that Ryan Rempel will do all in his power to support Mac OS X 10.5 on as much older, unsupported hardware as possible with the next version of XPostFacto.

Of course you have no direct input with Apple. However you are a guru-level personage with a widely-read website whose arguments can affect the attitudes of a large portion of the most loyal section of Apple’s user base. And how many of those people are also promoting Apple with friends, acquaintances, and employers through word-of-mouth because of their long and continuing positive experience with Macs? How many would have a sour taste in their mouth if Macs made as recently as five years ago were suddenly unable to “keep up with the pack”, as it were?

Whenever I am questioned for being adamant about going to great lengths to prepare for an unlikely event, I always try to point out that it is not the unlikeliness of the event that matters. Much more important are the potential consequences that occur after the event happens. Sure, an external hard drive is an extra expense. Sure, maintaining backups takes a bit of time out of your day. But what are the consequences if the unlikely event of a hard drive failure happens? Disastrous, unless you’ve prepared for it.

We’re worried about what happens if Apple unchecks that box in Xcode to include the instructions necessary to run OS X on G3 processors at all.

We aren’t worried about which machines Apple will “officially” support on paper. We’re worried about what happens if Apple unchecks that box in Xcode to include the instructions necessary to run OS X on G3 processors at all. As I pointed out before, such a decision by Apple will make tools like XPostFacto largely useless, unless you know something I don’t. Getting around some installer limitations is a walk in the park in comparison. After an event like that, how many users will feel less secure in their decision to buy a Mac? How many will be less likely to continue promoting Macs because of that lack of security? How many will never see new features right in front of their faces (running a bit slowly perhaps) that could prompt them to go out and buy a faster Mac, or recommend a new Mac to others?

The positive benefits to both us and Apple would be many-layered and long-lasting, acting in concert to promote continued growth of the brand. In contrast, the consequences of Apple dropping G3 support would be very negative not just for users but for Apple. A single disillusioned Mac user can potentially dissuade dozens from purchasing a Mac. To avoid these consequences we need to be united in showing Apple that G3 support is not optional. There are people at Apple who make these decisions. If they look out and see a community that says, “We demand G3 support,” they are much more likely to decide to support G3s. If they see a community that says, “Eh, we’d like G3 support, but those machines are old so it’s OK if they get dropped,” they will be more likely to simply drop G3s not just on paper but at the software level.

For every geek like us who is focused primarily on the latest and greatest stuff, there are 50 people who just want stuff that works and keeps working, and Apple computers do that.

Like it or not, you do have influence on the world you live in. If by strongly supporting compatibility with old Macs we can move Apple toward making future OS X versions even more capable of gracefully degrading but still functioning on old hardware, so much the better. They’ve already shown that they are capable of making versions of OS X that function exceptionally well on low-powered hardware (iPhone, Apple TV). For every geek like us who is focused primarily on the latest and greatest stuff, there are 50 people who just want stuff that works and keeps working, and Apple computers do that. Since computers are general purpose tools, part of the definition of “keeps working” should include continued compatibility with important tasks like email and browsing. Having artificial software limitations makes that very difficult. You get situations were the owner is fully aware that the machine could technically do the job and feels outraged at the software standing in his way for no good reason. If a person can look around his living room and see multiple Apple-branded devices that are still working, that’s what they will go out and purchase next time they are in the market for a similar device.

Macs that can stay current longer also have a reduced cost of ownership and can help to widen the user base beyond the upper middle classes who can afford to keep buying new ones every couple of years. Families that couldn’t afford two new Macs can pass on an old Mac and get one new one, and have them continue to function together smoothly. People who experience owning a Mac that keeps being useful for years feel better about saving up precious dollars to buy another one. Not to mention reducing the cost to schools, libraries and other low-budget organizations who often serve as silent but powerfully effective advertisements for the Mac platform. Many teachers, administrators, students and others have been prompted by their experiences to purchase their own Macs. Keeping these old Macs in circulation is not a net loss for Apple. Those who can afford new ones will continue to buy new ones and donate or sell their old hardware.

It is a huge win for Apple every time the IT staff at any organization can point to a bunch of Macs and say, “Look, they all still work and they’re incredibly easy to administrate because they can all run the same version of everything despite the fact that they are different models and different ages. Macs are reliable and save us a ton of money, so let’s get more.” And so far, it is still possible for an IT person to say that in many cases. That is a powerful argument that nontechnical business types will listen to.

I think G3 support is extremely important, and I wish you’d be more positive about promoting it as an absolute necessity, rather than making out like it would simply be a nice thing to have….

I could go on, but I think you get the point. I think G3 support is extremely important, and I wish you’d be more positive about promoting it as an absolute necessity, rather than making out like it would simply be a nice thing to have and tying it down with a whole range of qualifiers. Let them worry about which machines to “officially” support and which machines are “recommended”, they are good at that. Let us show them that unofficial support needs to continue to go much further back or else there will be a lot of very unhappy Mac loyalists. When the last G3 is at least 8 years old, it would be more acceptable to drop them, although 10+ years would be preferable. At that point those machines would be much closer to being actually useless, even for the specialized uses I’ve mentioned. Even the incredibly upgradeable Power Macs will be experiencing a lack of new PCI cards at that point. As it stands we have brand new PCI cards and CPU upgrades out the wazoo for almost any common computer technology.

That is my opinion. You are of course fully entitled to your own. Oh, did I say quick rejoinders? Sorry, I do get carried away at times. 8^)

Thanks,
Kris F.


 

Hi again, Kris,

We’re not that far apart. We have long maintained that it’s more work to take support out of OS X than to leave it in – whether that’s for ADB ports, SCSI, older Mac RS-422 serial ports, pre-AGP graphics, or G3 processors. Yes, there are some significant benefits of AltiVec support in G4 and G5 CPUs. Yes, there are some big advantages to the bandwidth of AGP video. And yes, there are people who will be perfectly content with less than perky performance.

…I’m one of those people who refuses to believe that Apple will break support for G3 processors.

I’m looking at official support, and I’m one of those people who refuses to believe that Apple will break support for G3 processors. I firmly believe that they will support all Macs up to five years old and many Macs up to eight years old officially – and that Leopard will be able to run on older, slower G3 Macs with pre-AGP video. There is no reason for Apple to remove support for PCI video, especially since it’s the only way to add additional video cards to G4 Power Macs that have AGP video.

I agree with all of your reasons for including G3 support in OS X 10.5, and the only reason I’m not “strident” in demanding it is that I believe it’s going to be there. Apple would be incredibly foolish to completely abandon the G3 at this point – and even more foolish to deliberately compile OS X in such a way that it can no longer run on G3 Macs.

My bigger concern is the premature abandonment of the entire range of PowerPC Macs in some future version of OS X. Some fear mongers are claiming that OS X 10.6 will have no PowerPC support, which I find ludicrous. That version, which probably won’t appear until sometime in 2009 or 2010, should still support the 2005 PowerBook G4, iBook G4, and G4 Mac mini models – as well as the Power Mac G5 Quad. Maybe in 2012 or so, when 10.7 might be expected, but not before that.

I can’t believe they would be so foolish as to block all G3 Macs, which means Leopard will be able to run on unsupported G3 Macs….

I guess the big difference is our expectations of Apple. I can’t believe they would be so foolish as to block all G3 Macs, which means Leopard will be able to run on unsupported G3 Macs, so I see no reason to petition Apple to retain G3 support. Instead, I’m trying to anticipate where Apple will draw the line for officially supported models this time around.

You seem to believe that Apple is likely to deliberately compile Leopard in such a way that is will not be able to run on a G3 Mac at all. Because of that belief, it’s important to bend Apple’s ear and convince the company that such a move would be a tragic mistake.

I’m sure all Low End Mac readers agree with both of us that Apple should not abandon G3 support, nor should they compile OS X to deliberately prevent it from working on G3 Macs until they reach the point of abandoning all PowerPC support.

Dan

Update: Apple did end up dropping all G3 support, and the installer won’t work with a G4 below 867 MHz, although there are a few different hacks that allow it to be installed on any Mac with a G4 CPU – even upgraded G3 models. Further, Apple moved OS X to an annual upgrade schedule starting with OS X 10.6 Snow Leopard, which is Intel only. OS X 10.7 Lion requires a 64-bit Core 2 Duo or later CPU and no longer supports software written only for PowerPC CPUs. OS X 10.8 Mountain Lion does not support 32-bit EFI, leaving out some Macs supported by Lion. OS X 10.9 and 10.10. run on any hardware that supports 10.8 Mountain Lion

Keywords: #osxleopard

searchword: g3andleopard