Are MacOS Support Periods Reasonable?

[redheeler]

Because of the amount of people I've seen on this forum trying to run a newer version of MacOS on unsupported Macs, I decided to do a quick analysis of whether or not Apple dropping support for a certain Mac model was justified, going back to 64-bit Mac models introduced in 2006.

I decided to set the threshold at which dropping support due to age would be acceptable at 9 years. Hardware of this age is becoming increasingly usable for basic tasks, and the interval at which many users upgrade their computer has become longer than it used to be. On top of that, the environmental benefit of being able to use older hardware for longer should be a source of pride for a company which claims to be environmentally-conscious. However, the burden of maintaining kexts and optimizing new versions of MacOS for the older hardware eventually exceeds the benefit it brings to the consumer, so it would be unreasonable to expect support to continue for a decade or more.

After the initial support period is up, security updates continue for another two years, though Apple tends to drop support with important apps like Xcode shortly after a new major version of MacOS is introduced. 11 years may seem like a long time to receive security updates, but it's not dissimilar from the two major competing platforms, Windows and Linux. Windows 7 will see almost 11 years of security updates, and that's assuming someone with 2009 hardware hasn't already updated to Windows 8 or 10. Current versions of Linux can likewise be run on almost any x86 PC or Mac from the past 13 years.

Below are the results of my analysis. TL;DR: MacOS support is usually dropped 1-2 years before it should be.

------​

Mac mini (Mid 2007)
MacBook (Mid 2007), MacBook (Late 2008), MacBook Air (Early 2008)

Logical Support Stoppage: OS X 10.9.5 (Mavericks)

Actual Support Stoppage: Mac OS X 10.7.5 (Lion) - 2 years short

Logical Reasons: Intel GMA graphics would likely result in subpar performance on Yosemite and later.

Actual Reasons: Apple chose to deprecate 32-bit kexts in OS X Mountain Lion and never released 64-bit versions for the GPUs in these Macs.

iMac (Late 2006), Mac Pro (Mid 2006), Mac Pro (Early 2007)
MacBook Pro (Late 2006)

Logical Support Stoppage: OS X 10.11.6 (El Capitan)

Actual Support Stoppage: Mac OS X 10.7.5 (Lion) - 4 years short

Logical Reasons: 32-bit EFI, old instruction set

Actual Reasons: Apple chose to deprecate 32-bit kexts in OS X Mountain Lion and never released 64-bit versions for the GPUs in these Macs.

iMac (Mid 2007)
MacBook Pro (Mid / Late 2007)

Logical Support Stoppage: OS X 10.11.6 (El Capitan)

Actual Support Stoppage: OS X 10.11.6 (El Capitan)

Logical Reasons: old instruction set, age (9 years)

Actual Reasons: old instruction set, age (9 years)

Mac mini (Early 2009), Mac mini (Late 2009), iMac (Mid 2008), iMac (Early 2009), iMac (Mid 2009), Mac Pro (Early 2008), Mac Pro (Early 2009)
MacBook (Aluminum Late 2008), MacBook (Early 2009), MacBook (Mid 2009), MacBook Air (Late 2008), MacBook Air (Mid 2009), MacBook Pro (Early 2008), MacBook Pro (Late 2008), MacBook Pro (Early 2009), MacBook Pro (Mid 2009)

Logical Support Stoppage: macOS 10.13.6 (High Sierra)

Actual Support Stoppage: OS X 10.11.6 (El Capitan) - 2 years short

Logical Reasons: age (9+ years)

Actual Reasons: age (7+ years)

Mac mini (Mid 2010), Mac mini (Mid 2011), iMac (Late 2009), iMac (Mid 2010), iMac (Mid 2011)
MacBook (Late 2009), MacBook (Mid 2010), MacBook Air (Late 2010), MacBook Air (Mid 2011), MacBook Pro (Mid 2010), MacBook Pro (Early 2011), MacBook Pro (Late 2011)

Logical Support Stoppage: macOS 10.14.6 (Mojave)

Actual Support Stoppage: macOS 10.13.6 (High Sierra) - 1 year short

Logical Reasons: outdated graphics (third-party Marzipan apps will expect Metal support)

Actual Reasons: Apple chose to deprecate OpenGL in MacOS Mojave, despite the fact there is nothing stopping the built-in Marzipan apps from working on OpenGL.

Mac Pro (Mid 2010), Mac Pro (Mid 2012)

Logical Support Stoppage: macOS 10.16.? (?)

Actual Support Stoppage: macOS 10.14.6 (Mojave) - 2 years short

Logical Reasons: old instruction set, age (9 years)

Actual Reasons: age (7 years)

 

[Longkeg]

Gee. I feel bad. All that exhaustive research and analysis. That long, well written post. And now, more than 24 hours later only 2 votes have been cast. And I don’t think those were the votes you were looking for. For what it’s worth Apple is going to do what Apple is going to do. Any “analysis” that doesn’t originate at 1 Infinite Loop is just so much noise as far as Cupertino is concerned.

 

[AidenShaw]

Any “analysis” that doesn’t originate at 1 Infinite Loop is just so much noise as far as Cupertino is concerned.

Wrong address.

 

[cyberdocwi]

Hello,

Thank you for taking a stab at this. I would like Apple to slow down these new releases of OS, or better yet, apply security patches to older OS, perhaps down to El Cap or even Mountain Lion. Yes, I know there is expense in keeping old software active, but there are people out there who enjoy operational computers, and don't wish to update them to a new OS, even if it is free. Send the security patch along, and let me get back to work.

I won't be moving my fleet beyond Sierra for a time to come. Sierra works. I don't have infinite RAM, and don't want to risk AFS yet, when HPFS has functioned well for me for years and years. I still use Microsoft Office 2011 Mac, as it works, is fast and reasonable, and does what I want it to do. I don't upgrade for the sake of upgrading. Yes I want to keep my systems patched, but I do have 32 bit code around, and not ready to purge it because someone in AppleLand wants me to.

Apple used to make awesome hardware. I don't like the new Macbook Pro's beyond 2015 -- the disruption of the keyboard, removal of even more ports, and the glued hard drive frankly pissed me off. I used to get excited about releases, and wanting even more performance, but Apple continues to arrogantly fail to consider the professional who doesn't want an army of dongles to loose, or who needs a function key instead of the CandyLand touchbar.

I use HDMI. I have several MagSafe power supplies hanging around. I need ports. I am a network administrator who uses ethernet, and cannot rely on WiFi.

Cyberdoc

 

[AustinIllini]

Hello,

Thank you for taking a stab at this. I would like Apple to slow down these new releases of OS, or better yet, apply security patches to older OS, perhaps down to El Cap or even Mountain Lion. Yes, I know there is expense in keeping old software active, but there are people out there who enjoy operational computers, and don't wish to update them to a new OS, even if it is free. Send the security patch along, and let me get back to work.

I won't be moving my fleet beyond Sierra for a time to come. Sierra works. I don't have infinite RAM, and don't want to risk AFS yet, when HPFS has functioned well for me for years and years. I still use Microsoft Office 2011 Mac, as it works, is fast and reasonable, and does what I want it to do. I don't upgrade for the sake of upgrading. Yes I want to keep my systems patched, but I do have 32 bit code around, and not ready to purge it because someone in AppleLand wants me to.

Apple used to make awesome hardware. I don't like the new Macbook Pro's beyond 2015 -- the disruption of the keyboard, removal of even more ports, and the glued hard drive frankly pissed me off. I used to get excited about releases, and wanting even more performance, but Apple continues to arrogantly fail to consider the professional who doesn't want an army of dongles to loose, or who needs a function key instead of the CandyLand touchbar.

I use HDMI. I have several MagSafe power supplies hanging around. I need ports. I am a network administrator who uses ethernet, and cannot rely on WiFi.

Cyberdoc

This type of attitude is the tech equivalent of anti-vaxxing. No one "upgrades for the sake of upgrading". You upgrade because the latest OS has improvements at the very least in security.

You upgrade because it's the responsible thing to do. Keep your tech and through it your data safe.

And you shouldn't notice the difference between HFS+ and APFS.

 

[jaw04005]

Reasonable? Maybe. But I don't think they should artificially stop supporting a machine in the new OS just because it's more than seven years old or whatever. Apple knows how many of each machine are still in use. They have App Store metrics that give them that information.

In a lot ways, Apple's push to treat Mac OS X like iOS has been positive. For example, unified code and features for end-users and developers. However, there's been side effects.

I work in the education field. There are many older modern iMacs (2010's) still in use in labs across our campus that can't be upgraded beyond El Capitan. There's some discussion this is due to a Wi-fi driver that's not 64-bit and the fact that this particular model doesn't have a Metal-capable GPU. These machines are very much capable devices. You're talking i5 and i7 machines with plenty of memory and SSDs. Honestly, they're still pretty darn fast machines.

This has caused major problems. Many of our teachers use iCloud and the iWork apps. If you open a document with the latest version of Pages on your iPad, you can no longer open it on your Mac because it's been "upgraded", etc. You end up having to go to iCloud.com and re-downloading a compatible copy to open on your Mac.

Unnecessary and annoying. And it's just another reason to use Google's web products for our classrooms even though Pages is a superior desktop publishing app.

I'd like to see Apple go back and do what they did with El Capitan. For the first time in a long time, they went back and updated older iMacs. They need to also fix iCloud to be more backward compatible and adjust for platform issues.

 

[velocityg4]

While I don't think they should keep having to release new versions of OS X for old models. As they would have to build and validate many of the drivers themselves. They shouldn't block old drivers or worse yet not allow third parties to release updated drivers (nVidia). If people want to risk a poor experience let them decide. Perhaps this hearkens back to G3 owners suing Apple because not all advertised OS X features were supported.

However, I do think they should provide security patches for deprecated versions of OS X as MS does with Windows for several years after updates are finished. It would make IT departments more accepting of Macs. Alternatively, they could only drop models every other version of OS X. Making those extended support versions.

 

[Tech198]

Everything must die eventually (... even humans do eventually)

The fact software can be "tuned' to run doesn't mean it should be supported.. As Apple doesn't cater for what the individual can do only... It's more about whats right for everyone without too much "bypassing"

By proving longer time for updates after an OS is no longer available may be shorter with Apple, but they could be doing that to force more people to upgrade as well.

After all, if you no longer have security updates, your more likely to look for updates apps, and get a newer version of the OS when you get a new mac, rather than the % of users who will find ways to keep it "as is"

This way, it gives Apple more credit when they boast in the media "how many users have using x version of our OS"

The other approach, wold be to just run virtual machines... (unless you have a really old Ma doesn't support Hyper-V)

 

[ScreenSavers]

I feel like some Macs have insanely long support periods and some dont... It should be more even. The 2006 Mac Pro was dropped too early, while the late-2007 MacBook Pros got a full 10 years of OS support. Some of the late 2009/2011 iMacs should've received more updates than they did (assuming the GPUs lasted.) The 2012 Mac Pro is very dissapointing.

 

[redheeler]

I feel like some Macs have insanely long support periods and some dont... It should be more even. The 2006 Mac Pro was dropped too early, while the late-2007 MacBook Pros got a full 10 years of OS support. Some of the late 2009/2011 iMacs should've received more updates than they did (assuming the GPUs lasted.) The 2012 Mac Pro is very dissapointing.

I agree. My Mid 2007 MacBook Pro is stuck at El Capitan and I am perfectly fine with that. 2011 MacBook Pros got two fewer years which makes little sense except from a forced obsolescence standpoint. They would (and can) run Mojave better than a Mid 2007 runs El Capitan.

 

[velocityg4]

I feel like some Macs have insanely long support periods and some dont... It should be more even. The 2006 Mac Pro was dropped too early, while the late-2007 MacBook Pros got a full 10 years of OS support. Some of the late 2009/2011 iMacs should've received more updates than they did (assuming the GPUs lasted.) The 2012 Mac Pro is very dissapointing.

Any of the Core Solo/Duo models got shafted in support life too. Same for 2008 Macbook buyers. The excuse of 32bit EFI was pretty weak. The CPU was 64-bit. They would have supported Mountain Lion and later just fine with drivers.

The 2012 Mac Pro isn't particularly surprising. It's just a 2010 Mac Pro. There was just a minor spec bump in 2012. The parts were still the same generation.

Apple has sometimes seemingly arbitrary tech cutoffs. Which should make any 2019 iMac customer weary. It is the only Mac without a T2 chip. I wouldn't be surprised if this fast tracks the iMac for being cut from OS support prematurely in macOS 10.17/18.

 

[lowendlinux]

Beyond support I honestly wish the world go back to a 3ish year release cycle and go back to charging for the OS.

Even better would be a stable vs unstable channel

It usually takes to a .4/5 release for the kinks to get worked out and by that time they’re ready to introduce the next shiny thing.

 

[z970]

This type of attitude is the tech equivalent of anti-vaxxing. No one "upgrades for the sake of upgrading". You upgrade because the latest OS has improvements at the very least in security.

You upgrade because it's the responsible thing to do. Keep your tech and through it your data safe.

And you shouldn't notice the difference between HFS+ and APFS.

Well that's an arrogant way to put it.

Whether it be unwanted foreign injections or unwanted software updates, it comes down to the user owning both their body and their personal computer, and exercising their own discretion on which road to take. That's a right nobody can hope to take away, no matter how enlightened they think they are by trying.

The responsible thing to do is maintain system functionality and reliability, even if that entails holding back from certain new releases. Otherwise, you have a "safe" new system that breaks a crucial part of your workflow, or is buggy in several areas and thus rendered unusable.

This is exactly why Apple needs to offer continuous security updates for older releases, even if they are accompanied by nothing else. It would at least be a gesture to the business crowd, who don't always jump to the newest version even if they're 3 years behind.

Right, HFS+ and APFS should be indistinguishable.

Until someone tries to file manage an APFS machine / server from 10.11 or older, and then discover that 10.11- doesn't know what an APFS is because Apple never bothered to update them with the necessary drivers in the first place...

 

[AustinIllini]

Well that's an arrogant way to put it.

Whether it be unwanted foreign injections or unwanted software updates, it comes down to the user owning both their body and their personal computer, and exercising their own discretion on which road to take. That's a right nobody can hope to take away, no matter how enlightened they think they are by trying.

The responsible thing to do is maintain system functionality and reliability, even if that entails holding back from certain new releases. Otherwise, you have a "safe" new system that breaks a crucial part of your workflow, or is buggy in several areas and thus rendered unusable.

This is exactly why Apple needs to offer continuous security updates for older releases, even if they are accompanied by nothing else. It would at least be a gesture to the business crowd, who don't always jump to the newest version even if they're 3 years behind.

Right, HFS+ and APFS should be indistinguishable.

Until someone tries to file manage an APFS machine / server from 10.11 or older, and then discover that 10.11- doesn't know what an APFS is because Apple never bothered to update them with the necessary drivers in the first place...

Well, it's reality.

And, once and for all, we need to clear this up. You own the hardware. Right to repair makes that clear.

HOWEVER, you do not own the software. You own a license to run the software on the phone. Apple can take that away.

Your opinion on responsibility is irrelevant. Apple only has to update the OSs they choose to update. It's your responsibility to keep your computer safe. If you misuse your computer, Apple can wash their hands of it.

 

[z970]

You own a license to run the software on the phone. Apple can take that away.

Apple only has to update the OSs they choose to update. It's your responsibility to keep your computer safe. If you misuse your computer, Apple can wash their hands of it.

True.

The hidden consequence of free upgrades.

 

[ApfelKuchen]

If the situation really was tick-tock 7 years from release, across the board, then it's a reasonable question. However, while there is a tick-tock time period for hardware support (Vintage/Obsolete), OS releases do not follow that strict clock - sometimes support ends sooner, sometimes later.

I don't have the expertise to know whether the "actual reasons" shown above are indeed the actual reasons. Far more likely (to me) is that the decisions were more complex than that. Example: "Apple chose to deprecate 32-bit kexts in OS X Mountain Lion and never released 64-bit versions for the GPUs in these Macs." Could Apple have written 64-bit kexts for those GPUs? It seems likely. Was it economically justifiable based on the number of in-service units vs. the engineering costs for development and testing? Were there other, unannounced deal-breaking dependencies? Were attempts made, but did they benchmark poorly? Unless someone leaks the contents of meetings and memos, how are we to know?

The fact that the lines drawn are so inconsistent from year-to-year, model-to-model strongly argues that these are not "you must buy new hardware" decisions. The free OS upgrades policy exists to strengthen the ecosystem; to keep as many users as possible, because the best way to grow the user base is to lose as few existing users along the way (the higher the loss rate, the more new users must be found in order to generate growth). Make new features available to as many users as practical (I originally wrote "possible" - but not everything that is possible is practical). Sometimes those new features are directly monetizable (Apple Music subscriptions), other times they are indirectly monetizable (customer satisfaction leading to future purchases of multiple products)...

But at some point, "unlimited customer satisfaction" becomes economically unjustifiable.

I suppose Apple could make this more cut-and-dried - a clear statement in terms and conditions "Every new Mac is eligible for annual OS upgrades for X years." It seems likely that making such a hard-and-fast promise would result in a shorter "X" than a longer "X," because the future is sufficiently unknowable that they'd have to err on the side of caution.

And as things stand... if I can expect that the latest OS will likely work on Macs that are up to 7 years old (and sometimes longer), that's really very good. It's reasonable to expect that a 3-year-old OS is likely to continue to work well, so that makes a ten-year practical life for the hardware.

 

[ScreenSavers]

I think the above post is very well said. Also, we are comparing it to Windows 10 in our heads, which can run on PCs from 2005. But the integration and quality control isn’t close. They will do anything they can to get people to install their software, whereas Apple is definitely pushing hardware.

 

[bunnspecial]

I don't get caught up strictly on the NUMBER of releases a given computer supports, but I too feel like the current 1-year release cycle is way too rapid.

As it is, it seems as though it's 3-6 months before all the kinks get worked out of a new OS from the time of the full blown public release. I've been running Catalina on one(secondary) computer since June, and I doubt that I would want to actually use it as my main OS until December or January(actually, I'll probably skip it, since I'm still fairly dependent on several 32 bit programs, or otherwise programs that themselves are 64 bit but have some 32 bit baggage tied to them).

It's often mentioned that computers like the last G5s and PowerBooks only received one OS update past the one they shipped with, but that doesn't seem so terrible in light of the fact that Leopard continued receiving security updates until mid-2011(6 years).

I'd just like 2 years to get to know an OS-and for Apple to really get it refined and the kinks ironed out-before it's replaced with something else. Doing that would naturally extend the OS support cycle, as it would also most likely turn into 4 years of security updates even after a system is deprecated from new releases rather than the current 2 years.

 

[redheeler]

If the situation really was tick-tock 7 years from release, across the board, then it's a reasonable question. However, while there is a tick-tock time period for hardware support (Vintage/Obsolete), OS releases do not follow that strict clock - sometimes support ends sooner, sometimes later.

I don't have the expertise to know whether the "actual reasons" shown above are indeed the actual reasons. Far more likely (to me) is that the decisions were more complex than that. Example: "Apple chose to deprecate 32-bit kexts in OS X Mountain Lion and never released 64-bit versions for the GPUs in these Macs." Could Apple have written 64-bit kexts for those GPUs? It seems likely. Was it economically justifiable based on the number of in-service units vs. the engineering costs for development and testing? Were there other, unannounced deal-breaking dependencies? Were attempts made, but did they benchmark poorly? Unless someone leaks the contents of meetings and memos, how are we to know?

The fact that the lines drawn are so inconsistent from year-to-year, model-to-model strongly argues that these are not "you must buy new hardware" decisions. The free OS upgrades policy exists to strengthen the ecosystem; to keep as many users as possible, because the best way to grow the user base is to lose as few existing users along the way (the higher the loss rate, the more new users must be found in order to generate growth). Make new features available to as many users as practical (I originally wrote "possible" - but not everything that is possible is practical). Sometimes those new features are directly monetizable (Apple Music subscriptions), other times they are indirectly monetizable (customer satisfaction leading to future purchases of multiple products)...

But at some point, "unlimited customer satisfaction" becomes economically unjustifiable.

I suppose Apple could make this more cut-and-dried - a clear statement in terms and conditions "Every new Mac is eligible for annual OS upgrades for X years." It seems likely that making such a hard-and-fast promise would result in a shorter "X" than a longer "X," because the future is sufficiently unknowable that they'd have to err on the side of caution.

And as things stand... if I can expect that the latest OS will likely work on Macs that are up to 7 years old (and sometimes longer), that's really very good. It's reasonable to expect that a 3-year-old OS is likely to continue to work well, so that makes a ten-year practical life for the hardware.

Apple's metrics for whether supporting a specific Mac model is economically justifiable mean nothing outside of Apple. Common sense will tell you they can be set arbitrarily to suite any agenda, as Apple can easily absorb the cost of supporting a handful of older Mac models.

Some Macs are supported for longer, and we can speculate regarding why that might be the case. But the fact remains, Apple has drawn the unofficial line at 7 years. That has been used consistently as a minimum cutoff since Sierra, though not as a maximum cutoff.