Callan Multibus 68k CPU board - Conclusion

I never would have thought it would keep me busy this long. It started with some talk about getting Unix edition 7 running on some computer, any computer. The idea was tempting at the time, but the temptation has now largely faded, but that is another story.

The Callan got dragged into this because I had the pieces of it in a box and I knew it used to run edition 7 unix on a mc68010. I thought, "let's cable up the hard drive to the original controller, fire it up, and write some software to read the blocks off the disk.

Given that 4 months have passed, you can guess that a lot of unexpected things took place. I am certain I have forgotten some or many of them. I'll make a list at the end of this diatribe, but first let me summarize how the whole thing has culminated.

I have recovered the entire contents of the 20M disk drive and have that preserved as a pair of tar files. This means that I could toss the hard drive, controller, and all the hardware and have what it (to my taste) the most interesting part of the whole thing.

I had a chance to remember many things about unix edition 7 from 1984 or so. No networking. You moved data using kermit over a serial port, or via 5 1/4 inch floppies. Filenames had 14 character limits. There was no /sbin directory (those things were in /etc). No symbolic links. No partition table on the disk.

Here is a list of some of the adventures along the way.

• Read and disassemble the boot roms.
• Write a python utility to download S records.
• Fix a libgcc problem in the gcc toolchain.
• Issues with multibus DMA and chassis.
• Nightmares with a bad power supply.
• Bad RAM card with shorted capacitor.
• Original HD controller will not read sectors.
• Use Gesswein MFM emulator to capture sector data.
• Extract disk image from that.
• Write extractor tool to read files from disk image.