> you'll sell units

Not just sell units, and probably not sell (many) more units, but you can sell units at a way higher profit margin. As pretty as the new Mac Pro is and as amazing as it is as a piece of technology, I have to believe that the one-time R&D costs of the Mac Pro are huge and directly contribute to the stagnated pace of upgrades.

While the cheese-grater Mac Pros weren't nearly as slick-looking as the oversized coffee cans, Apple could effectively throw out the whole internals of the cheese graters every year or two, put a mostly generic motherboard in it, with a wholly generic GPU and completely vanilla CPU/Memory/Disks and have a whoppingly fast machine with relatively minimal R&D cost, even while upgrading it in lock-step with external (intel/Nvidia/etc) evolutions. With the coffee cans, every single upgrade will require a huge redesign at a huge expense due to the tight integration.

While I understand the benefits of and necessity of tight integration in basically every single other product that Apple sells, its antithetical to the point of a Mac Pro.

Every OEM out there offers trash can form factor PCs these days that are better specced and cheaper than the Mac Pro. I think you overestimate the R&D costs that go into these things.

Just bite the bullet and hackintosh. It's ridiculously easy. I put one together in an afternoon with parts I had lying around. It even had GPU acceleration for all those silly UI animations to run smoothly. There are lists of hardware that are guaranteed to work. I'm not a Mac guy so I scrapped the system, but as a bored weekend project it was fun.

I had no issues installing updates, networking was fine etc.

It's not really an option in enterprise environment though. If you run a company and have to use Macs, then none of the options available at the moment give you access to lots of power and/or expandability.

Just built my second one because of their BS. Cost about $1100 for a computer that Apple would charge at least $5000 for. I've inspired friends as well because the process has become fairly easy nowadays.

I tried this a year or so ago, my only stumbling block was getting iMessage/iCloud working properly, which was very difficult, has that improved?

It's improved in that the causes of failure are now better understood. Getting it working means generating some UUIDs and things. e.g. https://www.tonymacx86.com/threads/an-idiots-guide-to-imessa...

No idea sorry. I put the system together to tinker with some iOS dev, but lost interest after a few months. I don't use any of the rest of the Mac ecosystem, so not sure how that functions.

What's special about iMessage/iCloud which poses an issue on a hackintosh?

Messages, Facetime and iCloud work for me on Sierra on my new hackintosh without any fiddling. On my previous Hackintosh, which I believe is running Mavericks, those services don't work.

I ran into this too. As far as I know, nope, it has not improved.

This is literally all I want. I'll even buy what will undoubtedly be an overpriced 8k monitor that they'll be selling along with it, instead of a new iMac.

I do not care what the box looks like.

Apple has trouble seeing themselves as sellers of commodity hardware like you describe - they never did, and there's a lot of pressure to innovate since Jobs left.

Due to this, they try to innovate on form-factor, which is the aspect they have most control over the product design. The problem is that this is directly at odds with 4) price, 5) expandability and (to a certain extent) easy to 7) spec-bump every year.

So I think it's very unlikely the next MacPro will fulfill this wish-list, even if it seems rational at first.

> Apple has trouble seeing themselves as sellers of commodity hardware like you describe

Switching from PowerPc to Intel chips didn't seem to make them much of a difference.

8. Be rackable (or at least, parallelepipedic) so the condition 5 holds.

I would buy two of these today.

That's exactly it.