A/UX and the Radius Rocket?

robin-fo

Well-known member
Hi everyone,

Another strange and silly question from me to the community:

Could you run A/UX inside a Radius Rocket virtual machine? This appears to be impossible because the Rocket uses single-partition bootdisks (like Mini vMac) while A/UX requires a SCSI disk with multiple partitions.

But.. Isn‘t there a SCSI board for the Rocket? So, assuming you could use real SCSI devices natively inside the VM, could you launch the Kernel and Root FS from there? You will certainly lack the Radius driver support and won‘t have support for the host volumes, but… why not??
 

cheesestraws

Well-known member
This sounds like a way to generate some really interesting crashes. Once my 950 is back in action, I might try it...
 

robin-fo

Well-known member
Yeah feel free to try @cheesestraws (another NuBus Mac will probably also do) 😉

My feeling is that the kernel will crash on booting like on an unsupported Mac or there will be no video output.. We will see!
 

Phipli

Well-known member
Yeah feel free to try @cheesestraws (another NuBus Mac will probably also do) 😉

My feeling is that the kernel will crash on booting like on an unsupported Mac or there will be no video output.. We will see!
The Rocket does pretend to be a Quadra 950, so... Perhaps.

But given how picky AUX is I'm sure something will cause an issue.

Interesting way to get AUX running on an 8100 though if it works 😆
 

mg.man

Well-known member
Or... (in a previous company) one of my clients using a Supercomputer to model landing a helicopter on the deck of a ship in 30ft seas...
 

robin-fo

Well-known member
Fun fact: You can launch A/UX on a Macintosh II without PMMU when using the Rocket as accelerator… but the Kernel will not take over the operation
 

adespoton

Well-known member
A/UX kernel expects certain fixed memory allocations, doesn't it? Are we talking 3.1.1 here, or an earlier version (2.x)?

It's possible we may have to get our fingers dirty with some kernel hacking to get this functional... but the result may mean a more usable A/UX less likely to unexplainably turtle if the hardware looks at it the wrong way.

There's no way I've got time to do anything but watch and comment, but once the SCSI cards are available, I'm more than happy to do that from the sidelines :)

I kind of feel it'd be interesting to get 0.7 at least loading the kernel to gain an understanding into what core parts are breaking before moving to a more complex kernel....
 
Top