Boot Camp freeware to dual boot Windows & MacOS is dead on all new ARM-core Macs

On 2020-06-24 13:41, Alan Browne wrote:

When was the last PPC Mac shipped anyway? A quick Google says Nov 2015.

Doh! 2005.

In article , Alan Browne
wrote:

It's not clear to me how that was stacked up (was it x86 code throughout
the parallels VM? Should be...). But it is clear that it worked.

an arm version of linux.

I wouldn't think so.

it is.

Parallels is an x86 gig.

more accurately, they're a hypervisor gig, using the hypervisor on x86.

part of the apple processors is also a hypervisor making it relatively
easy to virtualize another arm os.

So while it can be
Rosetta 2'd itself, it's made to host x86 OS'. Can Rosetta 2 do the
VM'd portion as well? Possibly. Indeed the icon for Mac 11
Virtualization has three heads ...

it might be possible to do that, but it would not work well.

Possibly Parallels and Apple worked together and Parallels have an ARM
version hosting ARM Linux.

yep.

In article , Alan Browne
wrote:


When was the last PPC Mac shipped anyway? A quick Google says Nov 2015.

i saw your correction to 2005, but that is wrong.

apple announced the intel transition in june, 2005, with the first
intel mac shipping in january, 2006.

the entire mac lineup transitioned to intel by august, 2006, at which
point apple stopped making powerpc macs.

existing stock was sold until it was gone, with refurb powerpc macs
selling for year or so after that.

several versions of the operating system fully supported both powerpc
and intel macs through 2011, six years after the transition was first
announced.

So while Apple can change CPU families and "manage the transition",
guess what they lose when they do that ? They lost *at least*
one customer... By making an orphan of my software collection,

Many developers (those worth it) transitioned to x86 painlessly enough.

for many developers, the transition was little more than a recompile.

some had to deal with endian issues because they chose not to write
endian-neutral code and it came back to bite them.

Some got mired in Carbon and refused to Cocoa. To be fair many had
valid business reasons to do so.

intel macs support both carbon and cocoa. many apps use both.

catalina (sep, 2019) is the first version to drop support of carbon
ahead of the arm transition which is also only cocoa. that's ~15 years
after the intel transition began.

Apple provided 3 OS rounds of Rosetta support as well. (Well 1 partial
round (Tiger), full round (Leopard) and optional in Snow Leopard).
Stretch SL out a few years and one was covered for a 4 - 5 year period.

lion (2011) was the first version to not include rosetta, which turned
out to be somewhat of a dud and many people skipped it, so it was
really mountain lion (2012) until powerpc apps could no longer be used
for most people, about 7 years from when the intel transition was first
announced.

but the story goes much deeper than that.

apple licensed rosetta from a company called transitive, which was
later bought by ibm, who had no interest in licensing it to apple or
anyone else.

https://www-03.ibm.com/press/us/en/pressrelease/26106.wss
ARMONK, NY - 18 Nov 2008: IBM (NYSE: IBM) today announced it
plans to acquire Transitive Corporation...
....
Transitive is a leader in cross-platform virtualization and a pioneer
in developing technologies that allow applications written for one
type of microprocessor and operating system to run on multiple
platforms -- with little or no modification. As a result, the
technology will enable customers to consolidate their Linux-based
applications onto the IBM systems that make the most sense for
their business needs.

rosetta 2 for arm is apple's own design and written in house, thus
*not* subject to the whims of another company.

it will last until there is no longer demand for it, at least 5 years.

In article , JF Mezei
wrote:

"*Boot Camp Is Dead On [all new ARM-based Macs]*"


Initially true. But if Microsoft starts to market ARM based Windows 10
that includes a Rosetta equivalent to run 8086 binaries, due boot might
return.

windows on arm already exists, although x86 emulation sucks.

Have there been any hints on what boot console will be used ? Will Apple
stick with EFI for its ARM based chips or use whatever it has developped
for ita Axx chips for iPhone/iPad/AppleTV ?

there's no reason to use efi.

In article , JF Mezei
wrote:


I've used both, and except in some very extreme edge cases, using a VM
is far, far better than using Boot Camp.


The VM in this case will have to be like Sheepshaver and provide a
platform emulation (complete with booting support).

no.

From what Apple provided in Monday, Rosetta is more of a utility to do a
one time convert of a file that contains 8086 binary to create another
file that contains ARM binary. (both allegedly packaged under the same
.APP directory structure to create that "Universal 2").

no.

There was mention of JIT file support, and I have to wonder if that too
will see the JIT executable fully received, converted and then launched.
(would need more details to confirm).

you always need more details, but that doesn't stop you from rambling.

In article , JF Mezei
wrote:


existing stock was sold until it was gone, with refurb powerpc macs
selling for year or so after that.

Suspect that discontinued models that are still in stock and don't sell
eventually go to "refurb" where Apple can price them low to get rid of
them without influencing published prices for new computers.

that doesn't change anything.

powerpc macs ceased being made in late 2006 and within a year or so,
all new and refurbs had been sold.

While Apple stated 2 year transition, I suspect that it will be much
faster, as was the case for the PowerpC to Intel transition as you
pointed out.

yep. a year from now, most, if not all macs, will be apple silicon,
except maybe the mac pro, which is a high end niche system.

For one thing, people will stop buying Macs now, awaiting the newer
ones. Secondly, Consider that the current lineup of Macs has seen recent
refreshes. So they are good to go without a new Intel model until the
ARM based ones come out.

sales will slow down, but they won't stop.

So it is in Apple's own interest to go wuickly on this.

The bigger question in my mind is how the first ARN chip will be
positioned. They produced the A12 based Mac Mini for developpers. Will
the first "real" one be a truly impressive high performance one for the
"Pro" models, or would it be an iPad class chip for the MacBook/Air and
normal iMacs?

the developer transition mac is for testing and not representative of
what future macs might be.



some had to deal with endian issues because they chose not to write
endian-neutral code and it came back to bite them.

In many cases, you have no choice because of what your code interfaces
with. Not all devices talk "text" in XML. And many older databases
stored data in binary to space space. Many config files are in binary to
prevent users editing them. Anything that is binary and stored in file
becomes endian sensitive.

that doesn't change the fact that some developers took shortcuts, which
came back to bite them in the ass.

apple licensed rosetta from a company called transitive, which was
later bought by ibm, who had no interest in licensing it to apple

I was puzzled by Apple's use of the erm "Rosetta". So did a big of googling.

"Rosetta" was not Transitive's (so not IBM). Apple Licenced
QuickTransit from Transitive and created its own brand "Rosetta" for its
implemnentation.

you're contradicting yourself.

But I suspect Rosetta2 is far more simpler as a file coverter as opposed
to converting machine code in-memory while it executes.

you suspect all sorts of things, very few of which have any relevance
to reality.

The file converter has the luxury of relinking the image and resolving
all exeternal references to point to the ARM versions of external
binaries. So when the translated app is launched, it truly launches as
a native app with all references to external subroutinesdynamic
libraries resolved to ARNM code right at launch time and no need to trap
any attempt to access untranslated code.

However, consider (whcih is now very rare and impossible on many
architvetures) a self modifying app that builds in memory some code to
which it then branches. Such an app would be translated and launch
normally, but when it writes to RAM code to which it intends to brand,
that code would still be x86 and it would then crash when it tries to
branch to it. (assuming it runs on a platform that allows one to branch
to data memory).

In am emulated environment, this would work because every instruction is
translated on the fly. However, the use of self modifying code is long
gone, except in virus attacks where buffer overloads attem,pt to insert
specific machine code that the attacker hopes the victim will branch to.
So by moving to Arm, it means those types of attacks will need to insert
ARM opcodes instead of 8086 ones.
#

you call that simpler????

In article , JF Mezei
wrote:


windows on arm already exists, although x86 emulation sucks.

Windows NT was also available on Alpha. and Windows was available on
Itanic as well.

win nt was available for power pc too.

https://en.wikipedia.org/wiki/Windows_NT_3.51#Overview
The release of Windows NT 3.51 was dubbed "the PowerPC release"
at Microsoft. The original intention was to release a PowerPC edition
of NT 3.5, but according to Microsoft's David Thompson, "we basically
sat around for 9 months fixing bugs while we waited for IBM to finish
the Power PC hardware". Editions of NT 3.51 were also released for
the x86, MIPS, and Alpha architectures

win nt has nothing to do with windows on arm today.


there's no reason to use efi.


EFI is already there , already written, no need to re-invent the wheel.

efi is *not* already there for apple silicon macs.

Rember that desktops require more boot functionality than a single drive
iphone, have wide variety of displays, graphic cards and need to be able
to boot from external disks of various interfaces (USB, thunderbolt etc(

that doesn't require efi.

On 2020-06-24 7:38 a.m., Arlen Holder wrote:
On Wed, 24 Jun 2020 10:24:50 -0400, nospam wrote:

apple didn't orphan anything.

Hi nospam,

As Paul just now eloquently exasperated...
o *In one swoop, _all_ your software no longer runs native on the ARM Mac.*

But if it runs just as well...

....why would you care?


This loss of this one freeware tool "could be" the harbinger of the demise
of freeware & 99-year licenses on ARM Mac.

What are you even talking about?


Time will tell on that prediction, but what we know for sure
with respect to software running native on the new ARM Mac...
o Apple effectively orphaned _everything_ (AFAICT) in one fell swoop!

Ummm.... ...no.

They didn't.


That means, if you want your software to run native on the ARM Mac
o You have to buy _new_ payware (which may likely be subscription based)

Ummmmm... ...no.


Worse...
o *You now have to buy payware to replace your freeware* (e.g., boot camp).

Ummmm... ...no.


In one swoop, _all_ your software no longer runs native on the ARM Mac.

Ummm... ...wrong.


While Apple has the lowest R&D % spend in all of high tech, never
underestimate the power of their MARKETING schemes, as Apple has a very
high profit margin indeed - and that's clearly not due to its R&D
expenditures.

On 2020-06-24 7:40 a.m., Arlen Holder wrote:
On Wed, 24 Jun 2020 14:38:11 -0000 (UTC), Arlen Holder wrote:

o *In one swoop, _all_ your software no longer runs native on the ARM Mac.*

I should have noted this was all your current "Mac" software (freeware and
payware) no longer runs native on the ARM Mac.

And I noted that there was no reason for you to care.


Where the point of this thread is the demise of Boot Camp freeware.
o The only native known alternative is payware to do the same thing.


VirtualBox from Oracle...

On 2020-06-24 12:55 p.m., JF Mezei wrote:
On 2020-06-24 00:45, Alan Baker wrote:

I've used both, and except in some very extreme edge cases, using a VM
is far, far better than using Boot Camp.


The VM in this case will have to be like Sheepshaver and provide a
platform emulation (complete with booting support).

Stop attempting to talk about things you don't understand.

Alan Baker wrote:
On 2020-06-24 7:38 a.m., Arlen Holder wrote:
On Wed, 24 Jun 2020 10:24:50 -0400, nospam wrote:

apple didn't orphan anything.

Hi nospam,

As Paul just now eloquently exasperated...
o *In one swoop, _all_ your software no longer runs native on the ARM
Mac.*

But if it runs just as well...

...why would you care?

If you've ever run heterogenous VMs, you'll know
what to expect.

They can run at 0.1x to 0.01x of the native clock speed.

VirtualBox in a homogenous ("untranslated") situation,
runs at around 0.9x. Giving you most of the CPU core you're
using. What VirtualBox doesn't always do a good job of,
is running multiple cores smoothly. There are some rough
edges under the hood. In some instances, it even rails
a core due to some sort of tasking issue.

Who wants to go from an Intel processor that turbos up
to 5GHz on a single thread, to some 2GHz ARM core doing
translation to make some legacy item run ?

I've been through all of this before. And have
no interest in "trying it" or anything else.

I used to run SoftWindows on PowerPC, as well
as VirtualPC (Connectix) back when it was its
own company.

Today, it's "native", "x86-on-x86", or "get outta town".
That's why I gave the "Magical" comment, not wanting
to waste the time explaining it.

Paul

On 2020-06-24 14:02:42 +0000, Paul said:
Alan Browne wrote:

MS' implementation of ARM Windows on the latest Surface Pro has been a
flop in several ways. Classic MS "get it out there...".

There was a Win10 objective though. To justify a Cloud orientation,
you have to cover all the users devices, from x86 desktops, laptops,
to ARM tablets or Windows Phone. They even have a version that
runs on the RPi (although graphics are gimped to a static image,
as the OS in that case was for "running the hardware" and not
intended as a replacement for a regular desktop).

I think Windows 10 was also offered as an installable option
on one of the Android smartphones.

Could I run a legacy win32 application in every case ?

Well, that's where the users aspirations come in.
Damn those users for expecting stuff to work for them.

*******

The transition for Apple, from x86 to ARM, is great for
their business plan, disruptive for their users. Take
me as an example. Sitting right next to me, right now,
is a PowerPC Mac, a G4. Now, look in the room. Do you
see an x86 Mac ? No. Why ? "Disruptive". How do I feel
today about an ARM version ? Are there enough adjectives ?

There are actually three Macintosh computers in this room.
What do they have ? All have PowerPC. Continuity. Continuity
helped make the sale.

So while Apple can change CPU families and "manage the transition",
guess what they lose when they do that ? They lost *at least*
one customer... By making an orphan of my software collection,
what else would you expect me to do ? There's not enough
reality distortion field for that.

Paul

Your software collection wasn't "orphaned". Apple and most developers
did an exceptional job at keeping most of it usable during the past CPU
changeovers, and all signs point to the Intel to Apple Silicon
chageover being equally as "smooth".

The change from PowerPC to Intel had Rosetta which translated apps
on-the-fly. The change from Intel to Apple Silicon will have Rosetta 2
which translates apps on installation and in some cases on-the-fly.

There was alos the Fat / Universal binaries which allowed apps to run
on both 68K Macs and PowerPC Macs, and then PowerPC Macs and Intel Macs

All of that means you could still use almost all your old apps until
you upgraded them. (There are of course always apps that don't work,
often due to the developers not following Apple's programming rules
and/or directly accessing the hardware.)

Of course, if you're happy with your current hardware and software
set-up, then there's no enforced requirement to update at all. I used a
G3 PowerMac for about 20 years until it died with a motherboard
failure. I only upgraded it to MacOS X because my useless ISP refused
to fix issues with their servers that meant MacOS 9 users could no
longer log onto the internet via dial-up conenctions.

On Wed, 24 Jun 2020 14:40:57 -0000 (UTC), Arlen Holder wrote:

The only native known alternative is payware to do the same thing.

Correction:

I should note that even the payware described in the news article for the
new Mac ARM, doesn't do the same thing as did the freeware on the old Mac
(they're quite different functionalities, actually).

Hence, in this case, until/unless a freeware solution is found for the new
Mac ARM, the article's lament appears to be valid that the new Mac ARM will
simply lose this freeware functionality that was ubiquitous on the old Mac.

Having had freeware dual boot capabilities on Windows/Linux for so long I
can't remember, I can't imagine that the Mac ARM users are happy losing
basic functionality that puts Mac ARM users back in the Stone Age of
computing without it.
--
A freeware solution for the Mac ARM will perhaps be found in the future.

On Wed, 24 Jun 2020 20:17:57 -0400, Paul wrote:

If you've ever run heterogenous VMs, you'll know
what to expect.

Hi Paul,

Anyone who suggests a VM to replace BootCamp/Grub, doesn't understand either.

As a purposefully helpful poster (unlike Alan Baker, who simply wants to
claim the Mac does everything better than does Windows), you need to
realize, if you don't already, that Alan Baker has an IQ of around 40 or 50
IMHO, so it's completely over his head that a VM isn't even close to the
functionality gained simply by "booting" to that 2nd operating system.

You'll have to explain that simple fact to him _multiple_ times...
o And, since he's an AppleSeed cultist, he _still_ won't get it.

Until/unless a freeware solution is found for the new Mac ARM, the
article's lament appears to be valid that the new Mac ARM will simply lose
this freeware functionality that was ubiquitous on the old Mac.

Having had freeware dual boot capabilities on Windows/Linux for so long I
can't remember, I can't imagine that the Mac ARM users are happy losing
basic functionality that puts Mac ARM users back in the Stone Age of
computing without it.

I've been through all of this before. And have
no interest in "trying it" or anything else.

Like you, and many others, I've been through the VM hassle before also.
o It's not even close to the functionality of booting directly to the OS.

As you're well aware, I've written tutorials on how to set up VMs in
Windows, and it's just horrid (particularly when you deal with the
hardware).

As such, like you, I'm done with classic VMs (and emulation also), which is
well known as I openly described why I'm done with VMs in detail on the
recent thread using WSL inside of Windows:
o *Tutorial for setting up Ubuntu as a Windows Subsystem for Linux WSL in Windows 10*
https://groups.google.com/forum/#!topic/alt.comp.freeware/rOT8xBWo9dk

Anyone who suggests a VM to replace BootCamp/Grub, doesn't understand either
--
A freeware solution for the Mac ARM will perhaps be found in the future.

On Thu, 25 Jun 2020 13:06:35 +1200, Your Name wrote:

Your software collection wasn't "orphaned". Apple and most developers
did an exceptional job at keeping most of it usable during the past CPU
changeovers, and all signs point to the Intel to Apple Silicon
chageover being equally as "smooth".

Hi Your Name,

*Nobody should be happy losing all their existing software functionality.*

I realize you're a well-known Type III apologist, so I simply remind you
that the topic is that the demise of the ability to multi boot on the new
Mac ARM is a pretty big hit in loss of basic boot functionality, which is
the topic of this thread after all.

*The new Mac ARM puts users back in the Stone Age of boot functionality.*

Given this loss of basic functionality on the new Mac ARM puts users in the
unenviable position of being slammed back to the Stone Age of computing,
Paul additionally rightly noted the loss of all prior Mac apps on the new
Mac ARM puts them even further backward in terms of their current
capabilities.

Since you're a well known apologist, none of those facts will have any
impact on you given that you justify Apple's actions in all cases, where
the scary _difference_ between Type III apologists is what scares me:
o Type I apologists === nospam: simply parrots Apple MARKETING always
o Type II apologists === Alan Browne: not malicious, just ill informed.
o Type III apologists === Your Name: actually _believes_ Apple MARKETING

Notice the key differences:
o Type I apologists don't believe a word they, themselves, say
o Type II apologists think for themselves but often filter out facts
o Type III apologists can't think for themselves - they believe the bull****.

The facts appear to be, whether you can comprehend them or not:
o While the Mac ARM may help Apple in _their_ brilliant business plans...
o The new Mac ARM instantly causes _loss_ of critical user functionality.

Only an apologists would be happy with that as the obvious 1st step.
--
Nobody should be happy losing all their existing software functionality.