PDA

View Full Version : How Convergent Technologies wrote their OS install/maintenance/software QIC tapes



firebirdta84
February 12th, 2015, 11:54 PM
I've recently reconnected with the owner of the only known surviving Convergent Technologies Factory CTIX installation and maintenance tapes for the MightyFrame. He's in Austria, and he's not comfortable sending his tapes to the USA, but he is willing to do things to try to read them.

He attempted to read one, and I have provided the files as a result, here: http://bit.ly/19e4HiF

Before I asked him to read any more of them, and risk further degradation through trial and error, I want to understand as fully as possible what I am looking at, so I can direct him to read them right on the first try.

There are multiple filemarks on these tapes, and I want to learn as much as I can about the logic and reason behind each of the files, and the filemarks that separate them.

So far, all I know is that 2 filemarks signify the end of the file set, or something like that. But, it's what's before those 2 filemarks I really seek to understand.

It seems that Al Kossow, as far as I know, did the latest work with these tapes. He hasn't found any surviving MightyFrame tapes that haven't been overwritten with something else.

So, for comparison, I have focused on Al's recovery of MiniFrame install tapes, which he provides on bitsavers here:
http://bitsavers.informatik.uni-stuttgart.de/bits/Convergent/miniframe/

Not all of the files are the same size, so I immediately ruled out that the filemarks are to keep the file size uniform.

It seems that some of these files are "trailers", as the bytes clearly show the repeated word "trailer" within, yet others are files with bytes...executables?

Al, I am hoping that you are able to chime in here on this thread, and may I take this opportunity to thank you for your dedicated work in this area, as well as everyone here how keeps this stuff alive. It is much appreciated!

Thanks always everyone,
-AJ

Al Kossow
February 13th, 2015, 10:09 AM
He attempted to read one, and I have provided the files as a result, here: http://bit.ly/19e4HiF


One of the problems with what he sent is there are clearly bad blocks, because you have fragments of what should be whole files.

ctfile5.dd is a kernel executable. I can see the trap table at the beginning and text strings that look like they are from the kernel
but then there are bits of the .text segment in ctfile6 and the symbol table in stfile7. There are bits of a ramdisk file system spread
over ctfile8-10

firebirdta84
February 13th, 2015, 10:39 PM
Thank you, Al...your comments here are extremely valuable and insightful to me.

So, so that I can establish a baseline in my understanding, can you provide some background on the f1, f2, f3...fa,fb files here, and so on?

http://bitsavers.informatik.uni-stuttgart.de/bits/Convergent/miniframe/

I'm interested in both how they were intended to be read by the original machine (in this case, the MiniFrame), as well as the exact command sequence on modern Linux that you used to extract them successfully.

Thanks!
-AJ