Hard disc image of RISCiX backup on archive.org

[iainfm]

I stumbled across this this other day https://archive.org/details/riscix and interest got the better of me.

It's a backup from around 1993. It has quite a lot of potentially interesting things on it, including early versions of bash and a couple of web servers. Plus (I think) sources for a lot of things (I've not looked closely yet). The domain in /etc suggests it's from Victoria University in Australia, and it has user accounts for 'ajw', 'alan' and 'alans'.

No idea if it's any use to anyone, but I've un-tarred it to a copy of the existing hdimage.hdf file (after wiping the root partition). It's not bootable (the /dev folder and goodness know what else would need to be built first), but it is mountable as a SCSI disc in Arculator. Just add the drive, point it at the image and mount it in RiX with

mount /dev/sd1a /mnt

The drive contents will then appear under /mnt.

Link to the hd image is: https://1drv.ms/u/s!AhjTXHcoiuP8gY4P21T ... A?e=SAj8US

Next thing I think I'll try it untarring it on top of a copy of the working hdimage.hdf disk and see if I can get it to boot...

[iainfm]

Did what I said I'd to and un-tarred the VU image on top of the hdimage.hdf and uploaded it here:

https://1drv.ms/u/s!AhjTXHcoiuP8gY4Q8cq ... A?e=e0KSRa

It's now bootable, but what it can do I don't know yet! (Edit: it breaks Xwindows for one!)

Only thing I've changed is the root password, which I've set to Tal540bo as per the other one.

[iainfm]

I've made the fresh, untarred version bootable. Just had to create the /dev folder, copy MAKEDEV into it, chroot to the mount point and run MAKEDEV.

Uploaded here: https://1drv.ms/u/s!AhjTXHcoiuP8gY4R12_ ... A?e=EIvIec

Root password has been reset to Tal540bo. No other changes made.

Happy huntings!

[awilliams]

I should have a look at that given as how 'ajw', 'alan' and 'alans' is unquestionably me.

[awilliams]

Well that's truly hysterical.

That *is* my machine the real one is on the bench behind me refusing to boot.

The 192.168.9.0/24 addresses are still my lan today, but it and a RISC PC are the only things surviving.

Other ip addresses in there show that it spent time at Melbourne Grammar School, its companions there were Netware 3 and Next (as dial up internet gateway), and also the outer eastern institute of technical and further education, now Swinburne University. These places are long after I would have done support for Victoria University who did have a lab of r140s. I had a site backwater.oeit.vic.edu.au where much scanned Acorn and associated documentation was available for quite a long time, but eventually Swinburne shut it down. Its possible that this image might have originated there since it fits the evidence in the hosts file, but I don't actually remember doing it.
Alan

[iainfm]

Hi Alan!

Wow, how amazing to find the original owner of the system!

I hope you get the real hardware running again - is it something you still use or has it been stored for a few years? If the latter, have a check inside for any battery corrosion!

Best wishes,
Iain

[awilliams]

This machine is just my A540+4M, its got two thin drives sandwiched into the drive bay. RISC iX would have been copied from an Acorn Melbourne demo R260 while I worked there.

It not been in storage but it doesn't get a lot of use, I used it last year as I finally got round to fixing my IO Podule for ARM3 by changing the 65C22 and its the last machine I have that podule would fit. Its also the last machine I have that is ARM and has an ADF10 socket.

Annoyingly it was working as well as its ever done which is to say randomly crashing has always been in its nature*, before I pulled it all apart and removed the battery & replaced it with an off board one. Its not booted again after that, so I think some track were only held together by the corrosion. It wasn't bad compared to some pix I have seen here but its still going to take some fixing so its on the rainy day pile for now.

Alan

*Its interesting that the Wikipedia page claims that early machines were 24MHz and later ones were 30Mhz which I think is completely the wrong way round. Mine was hand picked from early stock to ensure I got a 30MHz one. I think they lowered the frequency to increase the yield of usable chips. I have always suspected that this machines predisposition for unreliable behavior could be due to this. I suppose I could have just changed the crystal but that has only just occurred to me.

[paulb]

*Its interesting that the Wikipedia page claims that early machines were 24MHz and later ones were 30Mhz which I think is completely the wrong way round.

This being the RISC iX article. I see that this 18-year-old edit in Wikipedia's "write what I think I know" era introduced the claim, now expunged.

Mine was hand picked from early stock to ensure I got a 30MHz one. I think they lowered the frequency to increase the yield of usable chips. I have always suspected that this machines predisposition for unreliable behavior could be due to this. I suppose I could have just changed the crystal but that has only just occurred to me.

I think the service manual has different oscillator frequency configurations, but I didn't spend much time looking into it.

[AndyMc1280]

I don't know if my R260 is early or late but it came with a 26mhz arm3 and no FPA Socket.

[IanJeffray]

I don't know if my R260 is early or late but it came with a 26mhz arm3 and no FPA Socket.

IME, and from all I've seen, 26MHz was the stock speed for the standard plastic PGA ARM3 in R260/A540 and only the later rarer boards with QFP ARM3 and FPA socket were clocked higher.

[AndyMc1280]

I don't know if my R260 is early or late but it came with a 26mhz arm3 and no FPA Socket.

IME, and from all I've seen, 26MHz was the stock speed for the standard plastic PGA ARM3 in R260/A540 and only the later rarer boards with QFP ARM3 and FPA socket were clocked higher.

Well you got my old card

[awilliams]

I appear to be completely and inexplicably delusional.
It seems I have no actual fact to sustain any previous comments here about clock speeds.
I pulled the CPU out of my A540 (AKB50 SN 985) , which I have believed to be 30Meg for the entire time I have owned it only to find that its actually 26MHz. So erm oops, it seems everything I said above is totally wrong.

Alan

[paulb]

I pulled the CPU out of my A540 (AKB50 SN 985) , which I have believed to be 30Meg for the entire time I have owned it only to find that its actually 26MHz. So erm oops, it seems everything I said above is totally wrong.

Don't worry: the Wikipedia page didn't have any evidence for its claim, and I am starting to think that reviewed machines had 30MHz parts, only for Acorn to settle on slower parts in production machines due to yield, stability and FPA compatibility issues.

[IanJeffray]

reviewed machines had 30MHz parts, only for Acorn to settle on slower parts in production machines due to yield, stability and FPA compatibility issues.

26MHz was PGA, 30MHz/33MHz was QFP. Have you evidence for Acorn clocking the PGAs at 30MHz ?

[paulb]

26MHz was PGA, 30MHz/33MHz was QFP. Have you evidence for Acorn clocking the PGAs at 30MHz ?

Well, only the reviews of the R260 in PCW and Byte mention 30MHz parts. Acorn's brochure doesn't commit to any particular frequency.

[awilliams]

reviewed machines had 30MHz parts, only for Acorn to settle on slower parts in production machines due to yield, stability and FPA compatibility issues.

26MHz was PGA, 30MHz/33MHz was QFP. Have you evidence for Acorn clocking the PGAs at 30MHz ?

The very end of this page on crhis's acorns shows a very early r260 (#12) at 30mhz.

https://chrisacorns.computinghistory.or ... .html#R260
Also one on this page, (#145)
https://www.computinghistory.org.uk/det ... corn-R260/

Maybe the cpu grading was different between the A540 and the R260, with the better parts going to the R260s.
Maybe the advice I heard was get an early R260 not get an early A540, but I didn't understand there might be a build difference between two products other than cosmetic and drive contents.

[IanJeffray]

The very end of this page on crhis's acorns shows a very early r260 (#12) at 30mhz.

I respectfully suggest you don't take a photo of a single random sample board as being canonical of what Acorn actually shipped.
You may be right, but I don't consider this proof.