Apple's Path to Arm-Based Macs Could Start With a New 12-Inch MacBook

[Scottsdale]

.... you end up with a needless expensive product where no task will ever be able to utilize >50% of the power

Even if we were to ignore the engineering problems and having 2 CPUs sucking at the battery, to what point?

People seem to forget that Apple has everything under control this time. They can design an ARM chip in a way that is optimal for AMD64 (cos thats what the apps running under Catalina actually are) emulation.
They had years making sure XCode won't put anything into binaries that would turn out problematic.

The idea kinda reminds me of the Phase5 PPC cards for the Amiga which also featured a 680x0. Insane engineering but performance of the PPC part was crippled due to the 680x0 being on the same memory bus. In the end putting that effort into developing a 68k EMU would have been the much better alternative (as proven by even the first version of MorphOS being released a few years later).

The whole point is the Intel CPU would rarely turn on due to the hardware going into a sleep state. I earned a degree in computer engineering. It’s not the easiest, and it would cost Apple more, but the A-series SoC wouldn’t cost Apple much. And they could build reliance on the Apple ecosystem if done properly. It will take lots of engineering and they may mess it up, but if they have two working kernels, they can make it work. It would be awesome as far as seamless transition for the vast majority of users. And power users would have no problems switching now. Whereas they will likely wait years to ever switch until Apple can successfully emulate or all the professional apps can run on ARM. A lengthy headache for everyone but Apple. This makes the most sense for Apple developers, for users and for power/professional users. It’s a way to strategize and methodically build reliance on the A-series SoC while cutting out Intel in five years when people can actually have apps in place. Like Microsoft’s Surface Pro X. It’s amazing but it’s a glorified paperweight as 95% of the programs will not run or are too slow. Apple has such a better opportunity thanks to iOS and iPadOS. It could be a seamless transition for everyone except Apple. The A-series SoC costs Apple next to nothing, and Intel CPUs are incredibly expensive as the most expensive component in a Mac, but to look at it from anyone but Tim Cook’s perspective it makes far more sense to please the masses and make software support work and the seamless transition just work for everyone. The only company that would immediately be impacted would be Apple to build out the software engineering capabilities which we all know isn’t their strong suit. But this may be the thing that gets them motivated to do to better the products and Make Apple Great Again.
[automerge]1592236300[/automerge]

.... you end up with a needless expensive product where no task will ever be able to utilize >50% of the power

Even if we were to ignore the engineering problems and having 2 CPUs sucking at the battery, to what point?

People seem to forget that Apple has everything under control this time. They can design an ARM chip in a way that is optimal for AMD64 (cos thats what the apps running under Catalina actually are) emulation.
They had years making sure XCode won't put anything into binaries that would turn out problematic.

The idea kinda reminds me of the Phase5 PPC cards for the Amiga which also featured a 680x0. Insane engineering but performance of the PPC part was crippled due to the 680x0 being on the same memory bus. In the end putting that effort into developing a 68k EMU would have been the much better alternative (as proven by even the first version of MorphOS being released a few years later).

Not true at all. If done properly, from this computer engineer’s perspective, it would most of the time keep the Intel CPU in a suspended power state until the Intel CPU is needed. The A-series could run many SoCs before they ever got close to use as much power as the Intel CPU. If you know exactly what I am talking about you realize that only professional users need Intel most of the time and cannot emulate that well and end up with an exactly similar battery experience. About 90% of users get double or quadruple the battery life!!!
[automerge]1592236600[/automerge]

You can do that now. Run your Raspberry Pi (or AWS ARM instance) on a Remote Desktop on your x86 MacBook. Or run an x86 Mac mini (or AWS instance) on a Remote Desktop on some hypothetical ARM MacBook or on your iPad. Putting them in one box is a hardware and OS design nightmare (even just the T2 touchbar thingy causes OS problems).
[automerge]1592107191[/automerge]

No. The idea is to make a seamless experience whether it’s a basic user running Microsoft Office which would run the iPadOS version and get maybe 4x the battery life. Or, you would have a professional user running Adobe Premiere all the time and they would get the exact same abysmal battery time. I would guesstimate that 90% of Mac users would experience at least double if not quadruple the battery life as the Intel CPU would sit in there running not idle but in suspended power state until needed.

A lot of people say no without understanding the engineering behind it. As a computer engineer I say it’s 100% doable but Apple would need to up their game which they have to do anyways to emulate x86 on ARM. Microsoft has done it poorly. But the Surface Pro X gets amazing battery life. It’s just a dog at running any x86 app as the emulator is no good. If they had an Intel/AMD CPU running in a power suspended state until it needed they could make all users happy with that device.

 

[Gerdi]

So, eight Qualcomm Kryo 495 cores score 2900, and four Intel Sunny Cove cores score 2600? (Geekbench's browser actually lists 2909 for the Air i7.) Sounds like Intel is way ahead on that one?

You still seem not to understand. Ultimatively the only interesting metric is how much performance you get out of your system, not how many cores are under the hood. So no Intel is not ahead - they are behind.

And then of course, in a big.LITTLE setup, the small 4 cores only contribute less than 25% of overall performance. If we take the Winzip benchmark as example and only use the 4 large core we are still getting 14000MIPS - still beating Intel 10W and comsuming even less than 7W.

*Regarding 2909 for Air in Geekbench browser, i was using Windows results (not MacOS results) to make the systems comparable.

 

[57004]

I don't mind this idea. It would be nice to see some competition to x86/64.

However, I really hope Apple won't take this opportunity to create a more locked-down macOS as well, similar to iOS or Windows 10S. Locked down to the app store, and no more full filesystem access etc. That would totally kill the Mac platform for me.

 

[cmaier]

I don't mind this idea. It would be nice to see some competition to x86/64.

However, I really hope Apple won't take this opportunity to create a more locked-down macOS as well, similar to iOS or Windows 10S. Locked down to the app store, and no more full filesystem access etc. That would totally kill the Mac platform for me.

I don't think they will do that. If they ever get to the point where that makes sense for them, they'll just deprecate macOS and replace it with a variant of iPadOS with better window management.

 

[MikeZTM]

No. Intel x86 performance is still slightly faster if there are on the same node with Apple SoC.

First thing first: they are not on same node.
Let Intel doing same node first and then we are talking.

Apple 100% have more confidential Intel information than us, so they definitely know how "good" Intel's 10nm and 7nm node are.

CPU or SoC is a product. If someone cannot ship working product for a node then they can not claim they have any advantage if given the same node.

BTW we did see Intel's 10nm Skylake aka Cannonlake running hotter and slower than 14nm counterparts. I do not think just a better node will revive Intel.

 

[firewood]

As a computer engineer I say it’s 100% doable ...

"Doable" is rarely a good engineering metric. I've actually helped design working non-homogeneous multi-processor systems. Works for special purpose computing, but is a software and hardware nightmare for efficient general purpose computing. Better to do two separate boxes, and let them talk.
[automerge]1592243877[/automerge]

However, I really hope Apple won't take this opportunity to create a more locked-down macOS as well, similar to iOS or Windows 10S. Locked down to the app store, and no more full filesystem access etc. That would totally kill the Mac platform for me.

They can't lock things down so much that developers can't use some Apple product to develop and debug highly complex sophisticated apps, or they shoot their own App Store in the foot.

They could lock things down for non-(paid)-developers though. Extra revenue stream.

So they could kill it for you (if you aren't a dev), and replace you with more new and profitable customers.

 

[ph001bi]

Do you think that Apple just removed the PowerPC chip from the motherboard and soldered in an X86 chip ?

Of course there was a major hardware design change when Apple went from the PowerPC chip to an X86 chip.
[automerge]1592067742[/automerge]


I think that one of the major users of this would be students and you would definitely hear many complaints about not being able to game on it (other than iOS games).

When I said No design change I meant no external design change. Of course the motherboard was totally different.
[automerge]1592248018[/automerge]

That may be true now, but remember how quickly Apple dropped support for PPC machines? All the people who just bought PM G5 computers were probably not feeling too pleased. The 12" was, and still is, a great machine for those who don't need a ton of power.

Apple continued to provide updates and upgrades to the main operating system until August 28, 2009 when they shipped Snow Leopard, which was X86 only.
That's 3 years after they sold their last Mac with PowerPC, and the hardware was supported until August 7, 2013 when it reached "obsolete" status, 7 years after they sold their last PowerPC Mac.
I don't know if they'll replicate the same sort of timeline but I think it's pretty good.

 

[chucker23n1]

You still seem not to understand. Ultimatively the only interesting metric is how much performance you get out of your system, not how many cores are under the hood. So no Intel is not ahead - they are behind.

An eight-core score, for many use cases, is not a good metric on performance, especially on a device as low-end as the MacBook Air.

Most stuff people do on such a machine taxes one core, not all.

And then of course, in a big.LITTLE setup, the small 4 cores only contribute less than 25% of overall performance.

True. OTOH, Intel's Turbo Boost means a single core can clock higher in short bursts, which helps with a ton of relevant applications.

If we take the Winzip benchmark as example and only use the 4 large core we are still getting 14000MIPS - still beating Intel 10W and comsuming even less than 7W.

Come on. Where's your source for "even less than 7W"?

 

[dgdosen]

I don't think they will do that. If they ever get to the point where that makes sense for them, they'll just deprecate macOS and replace it with a variant of iPadOS with better window management.

I don't share your optimism here - Apple is increasingly making it more difficult to install and run 3rd party apps. My take is that they'll make vendors go through some 'new improved' AppStore - in the name of security and privacy... Where the only thing really getting secured are the gates of Apple's walled garden.

I hope I'm wrong...

 

[Scottsdale]

When I said No design change I meant no external design change. Of course the motherboard was totally different.
[automerge]1592248018[/automerge]

Apple continued to provide updates and upgrades to the main operating system until August 28, 2009 when they shipped Snow Leopard, which was X86 only.
That's 3 years after they sold their last Mac with PowerPC, and the hardware was supported until August 7, 2013 when it reached "obsolete" status, 7 years after they sold their last PowerPC Mac.
I don't know if they'll replicate the same sort of timeline but I think it's pretty good.

Three years for someone who just dropped $6k on a maxed out MacBook Pro 16” yesterday will suck at $2k annually for usage of an outdated computer system and obsolete within three years. Has to be among worst Apple investments ever for any Apple consumer.
[automerge]1592308235[/automerge]

I don't share your optimism here - Apple is increasingly making it more difficult to install and run 3rd party apps. My take is that they'll make vendors go through some 'new improved' AppStore - in the name of security and privacy... Where the only thing really getting secured are the gates of Apple's walled garden.

I hope I'm wrong...

I actually fear the same thing. They may try to monopolize it just as they have with iOS.

 

[Gerdi]

Come on. Where's your source for "even less than 7W"?

This is a logical consequence of the fact, that if 8 cores using 7W max and i am only using 4 cores, that the used power is less - as the big cores already running at max frequency in the 8 core use-case - contrary to Intel, there is no throttling here.
That is unless you think that the 4 low performance cores do not use any power at all...i am assuming that we are looking at at least 1W less.

 

[Gerdi]

An eight-core score, for many use cases, is not a good metric on performance, especially on a device as low-end as the MacBook Air.

Most stuff people do on such a machine taxes one core, not all.

Then why bother to put multiple cores in these machines at all? Because people expecting they can run either multiple applications at once or application transparently using multiple cores. These are the use-cases which are relevant for the 7W power limit and not the single core use-cases. You will also observe, that even simple web-browser use-cases using multiple threads, for instance in order to prefetch web-content.
So the discussion here is what return of investment the user can expect in terms of performance when investing 7W (or 10W for that matter) of power. And you can hardly explain to a user why his 10W machine is slower than the other users 7W machine.

Now coming back to the discussion at hand, I was just illustrating how a relatively old architecture like a Cortex-A76/Cortex-A55 machine (Surface Pro X) beats the latest x86 10 W machine in a lower power envelope. For reference i also put numbers for Intels latest and just released 10nm 7W solution - the Lakefield SoC. We are almost looking at 2x lower performance compared to the Surface Pro X here.
Now imagine how a more recent Apple A14x will beat the crap out of these Intel machines - remember we are not talking 5W phone chips here.

I would not be afraid at all - switching to ARM is a big opportunity for Apple of getting the performance crown in each single power segment - and i mean by a healthy margin.

 

[chucker23n1]

Then why bother to put multiple cores in these machines at all?

Oh come on.

Systems with two cores are a net win for almost anyone: just shove all the system background processes on the other core, and your focused app gets an entire core for itself.

For four cores, the benefit is already a fair amount smaller, but still there.

As the core count increases, the returns diminish, except for specialty applications that are parallelized well, such as AOT compiling (JIT compilers like Roslyn don't parallelize nearly as well), video encoding, etc.

That's why it's not that big a deal that the MacBook Air starts with just two cores (but I do recommend that people get the $100 mid-range quad-core option; it's just a good deal).

There's a difference between "is multi-core perf a good measure of general use patterns" and "are multiple cores useful". The latter: yes, they are (though the more you add, the less so). The former: well, it depends. You will be using those cores, you just won't be taxing them. That's why Intel has strategies like Turbo Boost where cores get disabled altogether, using the heat savings to instead clock the remaining cores higher. And it's why ARM has strategies like big.LITTLE and Apple's Fusion where not all cores are equal.

Because people expecting they can run either multiple applications at once or application transparently using multiple cores.

I think you'll find that the general audience of a MacBook Air doesn't really actively use multiple applications at once. It's why something like iPadOS is heavily full-screen-focused, why macOS eventually added full-screen mode, why a lot of Windows users maximize everything, etc.: multitasking, as a mental concept, just isn't that great or needed for most people.

These are the use-cases which are relevant for the 7W power limit and not the single core use-cases. You will also observe, that even simple web-browser use-cases using multiple threads, for instance in order to prefetch web-content.

Briefly, yes, but again, much of the computation then eventually ends up in the thread running JS (almost nothing uses JS's limited thread-like concepts).

There's a reason Apple is so focused on high single-thread performance.

So the discussion here is what return of investment the user can expect in terms of performance when investing 7W (or 10W for that matter) of power. And you can hardly explain to a user why his 10W machine is slower than the other users 7W machine.

Almost nobody cares about what the CPU TDP rating is. Apple won't even tell you what theirs is, and you still haven't given a Microsoft or Qualcomm source on theirs being 7W. Not that it really matters. Yes, it's interesting on a theoretical level if they can do better performance at lower wattage, but at the end of the day, the final product matters. It's just a small number on a spec sheet. And in this case, it's so small, they don't seem to publish it at all.

Now coming back to the discussion at hand, I was just illustrating how a relatively old architecture like a Cortex-A76/Cortex-A55 machine (Surface Pro X) beats the latest x86 10 W machine in a lower power envelope.

Sure.

Now imagine how a more recent Apple A14x will beat the crap out of these Intel machines - remember we are not talking 5W phone chips here.

I think Apple's chips will do better than Ice Lake, yes. Where things get interesting are questions like: what about Apple's vs. Intel's mid-term roadmap? And: if these Macs contain x86 emulation, how much better is Apple's chip, in order to offset the emulation overhead?

I would not be afraid at all - switching to ARM is a big opportunity for Apple of getting the performance crown in each single power segment - and i mean by a healthy margin.

Could be.

Personally, it's probably the moment I sigh and move on, because I need Windows virtualized at tolerable performance. Or I go to two computers, which I'd really rather not.

 

[Analog Kid]

I don't share your optimism here - Apple is increasingly making it more difficult to install and run 3rd party apps. My take is that they'll make vendors go through some 'new improved' AppStore - in the name of security and privacy... Where the only thing really getting secured are the gates of Apple's walled garden.

I hope I'm wrong...

I don’t share your despair, but I think the AppStore will play a part in the transition. My understanding is that the AppStore receives bytecode and Apple completes the compilation before download. If that’s true, then it would suggest that applications in the AppStore now are ready to go cross platform.

The question is how they handle applications that aren’t in the AppStore currently. Will they:

1. Force everything through the App Store
2. Provide backwards compatibility by emulating on the fly
3. Provide a fat-binary format and let Xcode compile to both instruction sets
4. Leave it on developers to sell ARM and Intel versions of their products

What isn't going to happen, and I'd be willing to wager heavily on this, is Intel and ARM chips executing userland apps side by side. There is simply no need to do that and all it does is kick the can down the road.

I'm guessing 1 or 3, and I think it could very well be 1. Every transition, Apple has used a different strategy. PowerPC emulated 68000, Fat binaries bridged Intel/PowerPC. They have a way of providing backwards compatibility already, I think they may very well double down on it.

It doesn't really bother me much. Whenever possible I purchase through the AppStore-- both the applications and the payment system are more secure. There are a few applications that I use which aren't available through that route. Most could be moved to the AppStore if the developers chose. The one that might be an exception is LittleSnitch, which I suspect runs afoul of some AppStore rules.

I'm not sure what it would mean to the open source community. Most open source projects are available as source, so if they keep the iOS loophole of being able to install your own self signed binaries, then that's one solution. I install some stuff through homebrew, which would probably die in an AppStore only world, but there's no reason the projects I use couldn't be made available through the AppStore.

There are legitimate philosophical concerns with limiting to the AppStore, but for me (and I suspect for most users) they aren't practical limitations.

 

[chucker23n1]

I don’t share your despair, but I think the AppStore will play a part in the transition. My understanding is that the AppStore receives bytecode and Apple completes the compilation before download.

You can send Bitcode to the App Store, but it's optional (I believe the only exception is watchOS, where all submitted apps are Bitcode). Hard to say how many apps use it, but even if they made it mandatory today, there would be years of apps that wouldn't run.

So, no, not really an option.

 

[filmbuff]

I think this whole debate comes down to "will it run OSX or iOS". If the hypothetical ARM laptop runs full fledged OSX in its current form there isn't much Apple can do to prevent 3rd party software from running. Even SIP can be disabled if it comes down to it. The big risk is that they run iPad OS or some modified version of that, or some hybrid that is really locked down. For me that would forever limit it to being a '3rd device' to play with but never a daily driver, I would switch to Windows before I accepted that.

 

[MikeZTM]

I think this whole debate comes down to "will it run OSX or iOS". If the hypothetical ARM laptop runs full fledged OSX in its current form there isn't much Apple can do to prevent 3rd party software from running. Even SIP can be disabled if it comes down to it. The big risk is that they run iPad OS or some modified version of that, or some hybrid that is really locked down. For me that would forever limit it to being a '3rd device' to play with but never a daily driver, I would switch to Windows before I accepted that.

They can fully lock down a Intel Mac with macOS just like how they are locking down iOS.
Encrypting the bootloader is a supported feature from both Intel and AMD for quite a long time.

There's no guarantee apple will not lock down macOS. It just not related to CPU and we hope they will not do that in the near future.
Since Xcode still need to be run on macOS and developer tools have to have higher access they probably need to keep the macOS open as a developer machine.

 

[chucker23n1]

I think this whole debate comes down to "will it run OSX or iOS".

By definition, it will run macOS.

If the hypothetical ARM laptop runs full fledged OSX in its current form there isn't much Apple can do to prevent 3rd party software from running.

Sure there is. Look no further than iOS for proof. macOS and iOS are very similar operating systems. One key difference is that iOS is more locked down.

Even SIP can be disabled if it comes down to it.

If they want to make SIP mandatory, they'll do just that.

 

[firewood]

Sure there is. Look no further than iOS for proof. macOS and iOS are very similar operating systems. One key difference is that iOS is more locked down.

Assuming that one rationale for the Mac to exist at all is to support developers, they can't lock it down or developers won't be able to build and debug code, and there will be no apps in the App Store.
[automerge]1592338011[/automerge]

The question is how they handle applications that aren’t in the AppStore currently. Will they:

1. Provide a fat-binary format and let Xcode compile to both instruction set

The current binary format for both iOS and macOS is already in fat-binary format. Just run the lipo utility to inspect. Some existing libraries already include both x86-64 and arm64 binaries (and 32-bit stuff as well) so they will run, as is, inside either iOS (ARM) or Mac (x86 Simulator) apps.

 

[chucker23n1]

Assuming that one rationale for the Mac to exist at all is to support developers, they can't lock it down or developers won't be able to build and debug code, and there will be no apps in the App Store.

Almost no developer needs to disable SIP. macOS can be locked down a fair bit further before it becomes untenable for developers. Heck, you could if you wanted to make iPadOS a development environment.

 

[sidewinder3000]

Thing is, the MacBook Air is already the ultralight.

There's certainly bigger problems facing Apple, but yes, I think paralysis of choice is absolutely one when a company offers too many similar products.

What's your elevator pitch on whether I should get an Intel-based ultralight or an ARM-based ultralight?



Well, there were other issues, like keyboard reliability. At the end of the day, I feel like the easiest recommendation was the Air: yeah, it has an old CPU and a non-Retina screen, but its keyboard is great, it has many ports, and it's hugely popular.

Which was kind of a shame. They fixed that. Let's not get back to those days.



It adds to the difficulty of making recommendations. Do you recommend the Air because it still has a more reliable keyboard, or do you recommend one of the newer MacBooks?

And, in a different way, that too is a line-up problem. Give us clear differentiation. If they put the butterfly in this new one, that almost makes sense to me. Make it way thinner and lighter even than the Air, and make it clear that its keyboard isn't the kind you get if you type a lot; it's the kind you get if you want maximum portability.
[automerge]1592208820[/automerge]


You're right that the 2x2 grid didn't last long, but I really don't think that was complicated. You got the Cube if you wanted a more stylish, less high-end variant of the Power Mac. You got the eMac if you wanted a cheap variant of the iMac (really, the eMac was the iMac G3 with G4-based internals). And so on.

I think it's much easier to decide between a Mac mini, eMac, iMac, and Power Mac than it was between a MacBook, MacBook Air, and 2-port MacBook Pro. All were around $1000-1500, all were laptops, all were thin and light, all had minor pros and cons. They occupied a very similar space.

Until I see how Apple will be implementing its transition to ARM, I wouldn’t be able to make any recommendation.

 

[dgdosen]

I don’t share your despair, but I think the AppStore will play a part in the transition.
...
It doesn't really bother me much. Whenever possible I purchase through the AppStore-- both the applications and the payment system are more secure. There are a few applications that I use which aren't available through that route. Most could be moved to the AppStore if the developers chose. The one that might be an exception is LittleSnitch, which I suspect runs afoul of some AppStore rules.

I'm not sure what it would mean to the open source community. Most open source projects are available as source, so if they keep the iOS loophole of being able to install your own self signed binaries, then that's one solution. I install some stuff through homebrew, which would probably die in an AppStore only world, but there's no reason the projects I use couldn't be made available through the AppStore.

There are legitimate philosophical concerns with limiting to the AppStore, but for me (and I suspect for most users) they aren't practical limitations.

What you're describing is EXACTLY what I despair.

Apple is getting sued by the EU, partially because they're picking and choosing who gets to play on the iOS App Store outside the scope of the 30% vigorish and that's a problem. (See https://www.macrumors.com/2020/06/16/apple-threatens-to-remove-hey-from-app-store/). It's becoming a bigger issue on iOS, and they've always had an app store. The app store was fine during 'the startup phase', but now Apple is trying to use it more like a cudgel. Now that the "App" market is more mature, the app store should drastically change, and I welcome the EU poking their noses in there.

I'd think the issue would be a much more bitter pill to swallow if Apple said to open sources devs: "You now have to go through a new App Store". They'll balk. I'll balk from ever using the device.

There are contributors to the kinds of tools/languages/libraries that I use and install via Homebrew (or some package manager on Linux) because they're Posix first, not macos. This issue is compounded by the notion that nowadays, I run these components in containers. What is Apple going to do to limit that?

Will 90% the world be happy if you have to go through an app store to install software on a mac? Probably, but all in all, you're going to stifle productivity and creativity... and lose the hearts of developers.

 

[cmaier]

Apple is getting sued by the EU,

False.

 

[Analog Kid]

You can send Bitcode to the App Store, but it's optional (I believe the only exception is watchOS, where all submitted apps are Bitcode). Hard to say how many apps use it, but even if they made it mandatory today, there would be years of apps that wouldn't run.

So, no, not really an option.

Still an option, but maybe not as seamless as I'd thought. Telling devs they need to resubmit to be compatible with the latest OS isn't beyond the pale-- and I'm not sure it would be the first time they've done something like that. If a dev can't be bothered to re-upload an intermediate product of an App they've already made, how much support is that app really getting?

What you're describing is EXACTLY what I despair.

Apple is getting sued by the EU, partially because they're picking and choosing who gets to play on the iOS App Store outside the scope of the 30% vigorish and that's a problem. (See https://www.macrumors.com/2020/06/16/apple-threatens-to-remove-hey-from-app-store/). It's becoming a bigger issue on iOS, and they've always had an app store. The app store was fine during 'the startup phase', but now Apple is trying to use it more like a cudgel. Now that the "App" market is more mature, the app store should drastically change, and I welcome the EU poking their noses in there.

I'd think the issue would be a much more bitter pill to swallow if Apple said to open sources devs: "You now have to go through a new App Store". They'll balk. I'll balk from ever using the device.

There are contributors to the kinds of tools/languages/libraries that I use and install via Homebrew (or some package manager on Linux) because they're Posix first, not macos. This issue is compounded by the notion that nowadays, I run these components in containers. What is Apple going to do to limit that?

Will 90% the world be happy if you have to go through an app store to install software on a mac? Probably, but all in all, you're going to stifle productivity and creativity... and lose the hearts of developers.

I don't think you're wrong to be unhappy, but I think you're overstating the implications. I'd imagine it's over 99% that will be just fine going to the AppStore for everything if the experience is smooth. I can't imagine more than 1 in a hundred Mac users install open source code or ever open the terminal. Being able to seamlessly transition from an old creaky Intel Mac to a sleek fancy ARM Mac would just reinforce what a great setup the AppStore is.

I get that people fear the big brother in Cupertino dictating what they can do and how, and there's a part of me that agrees. For now though, there's a bigger part that appreciates what the AppStore provides. I used to want to feel like I controlled every nuance of my computer's operation, but now I mostly just want an appliance I can depend on and leave the tinkering for Raspberry Pis...

 

[Detnator]

Always thought this was an obvious candidate since they discontinued the 12” MacBook.

I don’t know that they’d call a new thin and light laptop (12“ MacBook replacement) sporting an Apple-designed ARM processor a “MacBook“— hopefully not. I remember when they moved to Intel chips, they rebranded the most of Mac line. It went from iBook to MacBook, Powerbook to MacBook Pro, Powermac to Mac Pro. I think just the iMac and the Mac Mini didn’t get rebranded, no? For the most part it helped distinguish the new Intel machines from previous PowerPC processor machines.

Apple Book? Apple Book Pro? Don’t know how I feel about the names but Apple has been using that scheme in recent years, right? Apple Watch, Apple TV, Apple Pencil, etc.

My hope is they bring this out with 12” and maybe 14” variants - the 12 being similar to the previous one, perhaps smaller even if thinner bezels are possible, and the 14” still thinner and lighter than the current 13” MBA (even if the footprint has to be similar or slightly larger), and they call these two MBA (like the previous gen 11 and 13 inch MBAs) with prices similar to those - something like $999 and $1099 for base specs - or maybe $1099 and $1199

The MBP range should remain the higher priced higher performance machines, and for that it’s ok to start at $1500+

The MacBook (not pro or air) name should be saved for a future very entry level priced machine - say $799, maybe $899 at most. It doesn’t need the fancy tech required to make it super thin and light (that’s Air only) or the higher priced performance specs of the Pro. It needs Apple’s quality build etc. (it’s not a budget $300 laptop) but otherwise compromising everything else for the price conscious.

The basic no frills entry priced machine is MacBook (like the basic no frills iPad and iPhone). And then Air and Pro are more $ because of genuine features over that (fancy tech for super thin and light and/or high performance specs).

This was Steve’s range, up to about 2012 and made the most sense to me. The 2015 MB should have always been the new MBA.

Hopefully something like this is how they go, but... who knows. Apple’s products and services are pretty sweet. Their marketing and product positioning are a bit of a head scratcher sometimes. Lol.