It’s been an extremely busy week, and I haven’t had much of a chance to do anything blog-related at all. Instead, I’m often left staring at piles of stuff I’ve meant to make a mention of on the blog, but haven’t gotten around to it yet.
As part of the large load of stuff which was salvaged for me by my friend at uni, was this pair of OEM Microsoft MS-DOS Version 4.01 5.25″ DS/DD 360kB install floppies. Unfortunately, there was no box or manual with it.
The two disks were labelled Select and Install.The disks were labelled with disk number 05806 for Install and 05807 for Select. It is copyright dated 1981-1988 which makes it at least a year older than I am.
That would make it a little bit of a challenge to get the data off.
There wasn’t much distinguishing on the underside except for a heat-pressed number into the disk jackets.
Even the media envelopes were not notable.
As usual, I got out my Kryoflux and Panasonic 5.25″ drive combination for the challenge. Increasing the sector retries to 150 was enough to get it read eventually but not without making lots of horrible squeaking and scraping noises. The magnetic material wouldn’t be around for much longer by the sounds of it.
I tried to boot the floppies on VMWare Workstation 8, only to find that it always complained that the CPU halted due to some illegal operation. Using VirtualBox was enough to boot from one of the floppies, but installing was not an option as the prompt asked to switch to the Select floppy, to which I did, but the system didn’t recognize. From the Kryoflux read-out, it’s not possible to verify the disk as not having being written over or possibly corrupted in the past – so maybe it’s not forensic quality.
Part of the reason may be because the disks have the notch cut out, enabling write on the diskettes. At least I could see the version string banner for the OS and interestingly, it does display the date and time correctly … maybe that has something to do with VirtualBox.
Maybe MS-DOS v4.01 is buggy, maybe the emulation is to fault – only when I try to run it from “period correct” hardware will I be sure. The Kryoflux results seem to imply the disk has additional data in the header and might have been rewritten on a consumer drive, so the archival quality of these disks are called into question. However, disks of this age are very fragile and efforts must be made to prioritize their recovery – so I got this one out of the way as early as I could.