[acomjean]

That’s the popular narrative (I think Steve Jobs said that when he pulled all the clone licenses). The clone makers thought they were taking away Apple sales , but were expanding the Mac market.

https://en.m.wikipedia.org/wiki/Power_Computing_Corporation

As a Mac purchaser in the 90s (actually a clone, a power computing model) the 10% or so savings off real Apple hardware made a difference. macs /clones weren’t cheap, but having some other manufacturers at least make them made it seem like Apple was more viable and you were getting decent value for your money.

Of course having motorola making clones and the CPUs made for weirdness, as well as the company writing the os for clones making competing hardware. Apple wasn’t doing well and I think they were trying anything to survive.

Steve Jobs sent a Rolodex card and welcomed power computing users (Like me) to Apple when they pulled the plug.

https://www.macworld.com/article/2998000/clone-wars-when-the...

And Wikipedia has more:

https://en.m.wikipedia.org/wiki/Macintosh_clone#Licensed_Mac...

[chongli]

Apple’s problem at that time was their dreadful supply chain and glut of models. How were people supposed to know whether to buy a Performa 5200 (my first computer), or a Performa 6200, or a Power Mac 7200, or an 8500, or a 9500, or... one of the dozens of models in between all those?

I loved my Performa 5200 and I miss that computer like crazy, but Apple made a huge mistake with all of those models and they paid the price for it by having to write off huge amounts of old inventory.

It’s funny. Apple’s biggest weakness in the early 90’s has been transformed into their greatest strength today. Talk about learning a lesson!

[wtallis]

Arguably, Apple didn't just learn a lesson, but actually overcorrected. A majority of the complaints I've seen about Apple's computers in the x86 era and especially in the past several years have an underlying dissatisfaction with Apple simply not offering machines for various major market segments.

Apple's lack of a traditional consumer desktop PC means the Mac Pro is always getting complaints from consumers who don't understand the difference between the desktop and workstation market segments. The trashcan Mac Pro was a pretty niche product even among workstations. The MacBook Pro is less of a swiss army knife than it once was, and their other notebooks are even less willing to prioritize anything over being thin and light. Apple's never even attempted to entice gamers (much to their vocal dismay and derision), even though no other major PC vendor has cared as much about ensuring every machine has a decent GPU.

[chongli]

I’ve heard people complain for years that Apple ignores the gaming market but I’ve never heard a compelling argument for how Apple would benefit by catering to it. Gamers are a notoriously fickle bunch, always chasing the best price/performance ratios. Apple customers are notoriously loyal and willing to shell out a lot of money on high margin products. I think the overlap between these two groups must be tiny, given their contrasting natures.

The people who do love gaming and prefer Macs are probably better off buying a console. Those machines, of course, have no margin at all. I think this is the way Apple looks at it as well.

[chipotle_coyote]

While I think you're right, I think the OP is also right in that it might behoove Apple to have answers for at least a couple other market segments. The 16" MBP could be the laptop that has more ports besides USB-C, for instance, even if it adds a couple millimeters in thickness. And it'd be nice to see the Mac mini and the iMacs be made a little more user-modifiable: user-upgradable RAM and storage drives across all of them, for instance. (This is assuming it's possible to just plug in any internal SSD into a T2-enabled Mac, which I'm not 100% sure about, admittedly.)

I don't think Apple needs to go all in on competing with PC makers, by any stretch. But it would have been nice to see a lower-end version of the Mac Pro that started at, say, $2999. And that's not too crazy, given that the first Intel-based Mac Pro started at $2499. (Actually, you could get a build-to-order one with a slower CPU that started at $2199!)

[rbanffy]

> Mac mini and the iMacs be made a little more user-modifiable: user-upgradable RAM and storage drives across all of them, for instance

The new Mini and 27" iMacs can have their RAM upgraded. I believe in the case of the iMac it's possible (although terrifying) to upgrade storage as well.

I'd too love to see a MacPro that's not a luxury item but, with the massive precision milled shell and exquisite structure, that is unlikely to happen. I'm guessing a substantial part of the BoM is in that chassis.

OTOH, you can get a lot of workstation for a lot less with the Mini and iMac/iMac Pro. 18-core and 256 GB of RAM is a lot of workstation.

[chipotle_coyote]

I didn't know the new Mini still could have its RAM upgraded -- that's good. I'd still like to see the 21" iMacs have upgradable RAM, too, and to have the storage upgradable without the "although terrifying" qualifier. :)

You're probably right about the Mac Pro and its chassis; to go cheaper they'd have to be willing to go a little exotic. I suspect the Mac mini is supposed to be in that space. I suspect that rather than a cheaper Mac Pro, we'd be more likely[1] to see the Mini divided into "Mini" and "Mini Pro" lines -- I know the current model is more "Pro" than the previous generations, but they could be split in a way that lets them go down a bit more on the lower end and adds a few more options on the high end.

[1] More likely than a cheaper Mac Pro, but still not actually likely to happen. I know it's a bit of dreamcasting. :)

[beervirus]

Pre-2020 iMacs could have their internal storage upgraded, although it’s not for the faint of heart. From what I’ve read though, it’s no longer possible for the 2020 model with the T2 chip.

[chongli]

I’ll be interested to see what effect, if any, the move to ARM has on their Mac Pro line. It may be the case that the baseline model is so expensive due to the cost of Xeons. I don’t know for sure, though, so it’s a bit of a wait and see.

[hakfoo]

There's probably almost two sub-markets in gaming.

There's the crowd that ogles price-performance, that buys things like Ryzen 3600s and RX480s and 1050s. They're DIY builders.

Then you've got the splash-the-cash crowd, who buy machines from boutique shops and expensive prebuilds with $500 paint jobs on the case and "We have a GPU that's 2% faster and 40% more expensive". Apple could potentially build something in that niche.

However, their shift to home-built silicon is not going to do them any favours as long as most gaming takes place on Windows.

[rbanffy]

Apple is (and has always been, really) a systems company. They sell hardware that may or may not run other system software, but they sell it as a bundle. Interestingly enough, Jobs' move ending licensing of MacOS to third parties is a lesson learned by IBM a couple years earlier, not with their PC, but with their mainframes. Their mainframe OS is licensed in terms similar to Apple's macOS: it can't run on non-IBM hardware. IBM did this at the point they realized others, such as Amdahl, could build mainframes that ran IBM software faster and cheaper than IBM hardware. IBM would be giving up a sizable portion of their profits if they allowed that and thus they didn't. The last ancestor of z/OS you can run on non-IBM hardware is MVS 3.8j.

Controlling both hardware and software roadmaps allows IBM to use one as a force-multiplier for the other.

The death of the mainframe has been predicted since the 80's and doesn't seem to be any nearer now than it was then.

It'd be fair to point out that Jobs could have learned this lesson without IBM's help - as NeXT was failing, it wound down its hardware business to sell its OS and other software to run on other hardware and operating systems. At that point, it couldn't innovate like they did with their workstation with MO storage (which was a bad idea) or the built-in DSP (which was a good one). They would have to differentiate on software alone, something that's much, much more difficult to do.

[jasomill]

As far as I know, all z/OS ancestors were capable of running on non-IBM systems, and so was z/OS itself until it stopped supporting the 31-bit ESA/390 architecture which was the last IBM mainframe architecture successfully cloned by any of the plug-compatible manufacturers.

MVS 3.8j was, however, the last version of MVS that was both available from IBM without a license and which lacked copyright protection under which IBM could prohibit unrestricted use.

And, to be perfectly pedantic, a number of z/Architecture "clones" now exist in the form of software emulators — including at least one[1] developed and sold by IBM itself — and these are perfectly capable of running more recent z/OS releases, possible license complications notwithstanding.

[1] https://www.ibm.com/products/z-development-test-environment

[rbanffy]

True. If you start today, you are limited to 3.8j, because IBM won't license z/OS to you. If you already have a license, you can continue using your software on non-IBM hardware, but I think IBM explicitly limits your ability to run z/OS the same way Apple does macOS.

As for the ZDTE, we can call it "virtual IBM hardware" ;-)

[shrubble]

The difference however was that IBM was operating under a consent decree that required them to license their OS under reasonable terms to others.

[karlshea]

I lusted after Power Computing's stuff back then, and their marketing was so great. I just ate up the "Let's kick Intel's ass" underdog vibe.

Of course then a couple years later all the new Macs were Intel.