We all know in the last change from PPC to Intel, eventually Rosetta and Universal Binary support was stopped. That's obvious as it exists as a stop gap transition to allow people to smoothly from all PPC to all intel.
But now we have the intel to Apple SoC transition. How long do you think Apple will give us all to move to a fully native Apple SoC environment?
This is substantively an "Apples to Oranges" comparison. Apple didn't own Rosetta. It was another company's technology that Apple licensed ( very likely had to pay a fee for every copy shipped). The length of time there was very intertwined with money and costs to continue to do so.
This Rosetta 2 works substantially different. For the most part is basically a complier that Apple developed to churn out a ARM binary from an x86-64 one. Besides any AMD/Intel patent it might get entangled in ttere isn't any licensing to be done. Once it is working then Apple can put it in maintenance mode where they make just about zero changes and the costs dramatic drop down much closer to zero per copy distributed.
Apple pretty likely won't have Rosetta 2 on the same timeline if it is something they have already paid for. There is lots of mostly comatose software that Apple bundled for more than several years with macOS. The software raid system. Apple File services has been basically comatose for many years. etc. etc. That is more the yard stick should be measuring against. Not Rosetta ( Transitive owned circa 2006 and IBM owned own. IBM bought the underyling tech in 2009 and put it on a path that was divergent from Apple's main focus. Plus IBM wasn't going to get browbeat on licensing terms as a "small company" with limited resources. )
Similarly, dropping the Intel systems is likely going to take longer this time because the installed base inertia is much higher now ( 100+ Million Intel Macs out there and most of them aren't going away any time soon.).
Also how will the first generation of Apple Mac SoC fare? I remember the first gen Core Duo was not supported that long by Apple and shortly after everything required at least the Core 2 Duo. Will the 2nd gen of Apple Mac SoC quickly make the 1st gen obsolete? This is basically a how will the early adopters fare?
some of that was getting caught up in the 32-bit support. There were Core Solo Macs too ( that didn't even have a 64 options).
This first set is not going to run into any 32-bit legacy stuff as Apple basically has chunked that out the windows the year before the transition regardless of instruction set. Apple has flushed substantially more stuff beforehand this time. [ Pretty ggod chance have worked on the transition process longer. But will transition a bit slower as Apple doesn't have a complete line up of processors to transition to. Laptop space ? yes. High end desktop? No. ]
Also as an aside Craig mentioned a new virtualisation app. Do you think we'll be able to virtualise other OS there like older Mac OS versions and Windows there also? I ask this because many people use bootcamp on a daily basis for native Windows on their Mac. A move to Apple SoC I think would kill bootcamp.
Supported (by Apple) older Mac OS versions. Very likely not. Someone comes up with spit-and-bailing wire-and-ducktape macOS x86 on QEMU probably will pop up (
https://wiki.qemu.org/Hosts/Mac ). Similar to Sheepshaver for running old mac OS 9 images. IMHO seems doubtful anyone like VMWare or Parallels is going to bother with it.
Probably dead. If there is no UEFI on the new Apple Silicon Macs ... then even more highly probably dead. Apple has walked away from very small user constituencies in the past.
These questions are making me put off the iMac purchase I was considering. I feel there's almost no point purchasing any intel Mac at this point in time. Still I feel these questions are really important to consider.
If booting 'raw' into Windows is some deep seated, hefty (performance and/or quirky compatibility I/O port ) requirement then buying an Intel Mac probably should be considered now if have substantive needs now. That ability is likely going away.
The length of time of the Intel->ARM transition probably is not going to be the same timeline as the PPC->intel one. Intel had a full line up of processors if Apple came over or not. They had one before Apple came, during when Apple used them ,and will still have one after Apple leaves. Apple doesn't. It is going to take Apple much more time to roll out a complete line up. And until they do that will extend the time that they'll have to actively support x86-64 macOS. Throw on top the inertia is much larger now and that also likely leads to an more extended time line. And the cherry on top is that Apple gets money very services ( subscription software whether on Intel or ARM mac. ) . Apple of 2020 doesn't get its money from the same places as Apple of 2005-6.
The folks talking "complete doom" for Intel macOS in three years are pretty likely wrong. It isn't going to be 10 years because even if there was zero instruction set transition don't get 10 years anyway. 4-5 years from now is likely range of last macOS and useful working lifetime after that is longer ( unless have super requirement for most bleeding edge macOS).
