Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

khronokernel.com

Post reply on HN
Print view View original

181–190 of 191 posts

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#181

Post by gambiting »

saagarjha wrote: post #174
gambiting wrote: post #171

Earlier quoted context omitted.

We use Unreal Build Accelerator which injects a custom dlib into clang to intercept the compilation process and distribute it to worker machines.

I’m curious why this needs to be code injection and not, like, a shell script?

Because...it's official behaviour that is fully supported by clang? If you want to add a hook on compilation start, it's literally the documented way - you include your own dlib with necessary overrides and then you can call your own methods at each compilation step. Not even sure how you'd do it with a shell script? You need to have knowledge of all the compilation and linking units, which....you have from within Clang.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#182

Post by egorfine »

benoau wrote: post #16
kylec wrote: post #2

This is a very silly restriction, at least to apply uniformly to all Macs. I think if you buy a more powerful Mac they should let you virtualize more Mac instances. Like an M5 maybe limit to 2, but maybe let an M5 Pro do 4 and an M5 Max do 8 or something.

Why should they impose a limit at all? Your hardware is a natural limit, you'll stop of your own accord when you reach its thresholds.

> Your hardware

They see it a bit differently.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#183

Post by matheusmoreira »

AussieWog93 wrote: post #88
array_key_first wrote: post #82

Earlier quoted context omitted.

All operating systems have bugs, and Apple doesn't have the QA it used to have. MacOS has basically been exclusively trending down in quality for a while now, while Linux continues to get better. What you have to realize is that what Linux distros are doing is inherently more complicated. They're making a general purpose operating system intended to run on every computer. Apple is making one operating system intended…

You're not wrong about the downwards trend in quality but we're still a long ways off from macOS or even Windows having the same level of QA issues that Linux does, on a regular desktop system.

The Windows computer I have to use takes 17 seconds to open the "updated" calculator. The old notepad opens instantly.

17 seconds. I timed it.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#184

Post by icedchai »

morphle wrote: post #119
icedchai wrote: post #94

Earlier quoted context omitted.

Are you talking about A/UX? That was one of the first Unix systems I was exposed to.

Yes but I had others too. BSD on both 68000 and PowerPC

Did you ever experiment with Mach Ten ? I'm thinking of putting it on an old Mac.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#185

Post by morphle »

icedchai wrote: post #184
morphle wrote: post #119

Earlier quoted context omitted.

Yes but I had others too. BSD on both 68000 and PowerPC

Did you ever experiment with Mach Ten ? I'm thinking of putting it on an old Mac.

Yes, I ran it also on 68000 and PowerPC Macs. I preferred MacOS with all the MPW environment and tools on top, the GUI was much better: a full WYSIWYG text editor that also was the command line, so you could compose text, copy and paste and also execute it. But that was invented with the workspace in Smalltalk-76 and recreated with MPW.

Email me if you need help restoring it on your Mac, or if you need parts to revive your hardware. I have at least one of every Mac since 1982 (yes I know the Lisa was introduced in januari 1983) including all floppies, CD-ROMs, books, screens, keyboards mice, Appletalk. Although some parts have rusted or decayed beyond repair. I hope someday somebody will buy the whole museum from me.

The best quality Unix we ran was BSDi, you'll find some of that still in NetBSD, OpenBSD and maybe FreeBSD.

The coolest Unix was IRIX though, but that was because of the Grafix code, not the Unix kernel.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#186

Post by john_strinlai »

matheusmoreira wrote: post #183
AussieWog93 wrote: post #88

Earlier quoted context omitted.

You're not wrong about the downwards trend in quality but we're still a long ways off from macOS or even Windows having the same level of QA issues that Linux does, on a regular desktop system.

The Windows computer I have to use takes 17 seconds to open the "updated" calculator. The old notepad opens instantly. 17 seconds. I timed it.

mine opens in <1 second. i wasnt even able to time it.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#187

Post by thealistra »

leptons wrote: post #124
VanTheBrand wrote: post #96

Earlier quoted context omitted.

The option is you have to buy another machine. There are mac ec2 instances and several mac cloud hosts that all would abuse this if they could, instead to stay compliant they buy more machines.

I tried to launch a MacOS instance on EC2 recently (on my work account), and was blocked. So I asked the IT dept and they said it's stupidly expensive to run a MacOS instance on EC2, and that they would just send me a Macbook Pro instead . I wish I were kidding.

You should be happy that you have such a financially aware IT dept. The machines are truly expensive and after managing a farm of 150 Mac minis at work I can tell why.

It is like 3 days running the EC2 buys you a Mac mini? And you can only rent the machine in day increments

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#188

Post by saagarjha »

gambiting wrote: post #181
saagarjha wrote: post #174

Earlier quoted context omitted.

I’m curious why this needs to be code injection and not, like, a shell script?

Because...it's official behaviour that is fully supported by clang? If you want to add a hook on compilation start, it's literally the documented way - you include your own dlib with necessary overrides and then you can call your own methods at each compilation step. Not even sure how you'd do it with a shell script? You need to have knowledge of all the compilation and linking units, which....you have from within Cl…

What's the interface this uses? I don't think I am familiar with it

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#189

Post by naikrovek »

stephenr wrote: post #177
naikrovek wrote: post #175

Earlier quoted context omitted.

Yes we all realize that. It’s MacOS VMs that we want to run.

Maybe I should have used the same dismissive tone. Imagine thinking everyone who buys a Mac and runs VMs wants to run heaps of macOS VMs.

why else would you buy a mac to run VMs?

arm64 hardware is cheap, x64 hardware is cheap and both of those can run as many Linux or Windows VMs as you have RAM to run.

Re: Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

#190

Post by gambiting »

saagarjha wrote: post #188
gambiting wrote: post #181

Earlier quoted context omitted.

Because...it's official behaviour that is fully supported by clang? If you want to add a hook on compilation start, it's literally the documented way - you include your own dlib with necessary overrides and then you can call your own methods at each compilation step. Not even sure how you'd do it with a shell script? You need to have knowledge of all the compilation and linking units, which....you have from within Cl…

What's the interface this uses? I don't think I am familiar with it

There's a very good explanation in this presentation :-)

https://static.linaro.org/connect/yvr18/presentations/yvr18-...

Post reply on HN