LOS 1.1B (British English) (recovery in progress)

[fri0701]

Exciting news!

I recently purchased a pair of Twiggy operating system disks and matching manual for Lisa Office System 1.1B - the British English localization. The disks were very moldy upon first inspection, and I was worried the information on them would be lost, as is the case with so many other Twiggies nowadays. However, after some careful cleaning, and with much help from Tim Colegrove and the use of his newly-restored Lisa 1 at The BYTE Shop, we were able to successfully* archive both disks. As far as I'm aware, this not only represents the first time a version of LOS 1.1 has been archived, but also the first time a British localization of LOS has as well.

*While the disks ended up being mostly readable, each disk had a few sectors that did not pass BLU's checksum test. We were able to get past these sectors by using the "Read ignoring checksum" option. The sectors in question are noted in the attached log files.
- Several sectors were not being used by any file, as reported by lisafsh-tool, and I have ignored modifying these.
- Several sectors were in files that I see are identical to the analogous file in LOS 1.0 except for the specific broken sectors, so I have assumed these are fully identical and have patched these disks with the information from a known good LOS 1.0 disk.
- Several sectors were, unfortunately, used by files that are unique to LOS 1.1, including the OS and Office System shell files. These sectors reported a checksum failure, but produced the same exact results across multiple reads of the disk. Since I noticed other cases where "Read ignoring checksum" produced random data each read, I have no choice for now but to assume that these sectors have at least some level of stability, and I've kept them as is. If another LOS 1.1 disk is archived, it would be certainly possible to compare the two images and, if necessary, patch a fix into these images.

So, the million dollar question - What's different about LOS 1.1B compared to LOS 1.0?
I haven't had the chance to actually install the system yet, so my analysis is somewhat limited to examining the disk images directly. As far as I've seen, the only differences in the phrase files are references to the "Apple-period" key combo, which have been replaced with "Apple-full stop". Surprisingly, the OS itself seems to be entirely unique, and is not just a relabeled LOS 1.0 or 1.2.

I'm posting the patched disks here, as well as a log of the disk errors we encountered. I'd very much enjoy seeing these run on someone's Lisa 1, if you're willing to attempt an install! If it does install correctly, a clean ProFile image would also be wonderful to have in the archives.

As always, if anyone has any Lisa software that hasn't been archived yet, please let me know. I don't want to speak for Tim, but I think it's safe to say we'd be more than happy to assist in the archiving of any Twiggy media, especially LisaTerminal, which is the current number 1 priority.

[rayarachelian]

Wow, this is a pretty unique find. Awesome.

[stepleton]

Wow! Do you know the story of where these disks came from?

It's a busy time for me, but I'll have to give these a shot when I can.

For the files which were damaged, like the shell files, I wonder if more analysis is possible. Could you run a shell file through a 68000 disassembler and identify which portion of the file came from the damaged sector? If we're very lucky, it may be possible to pinpoint the flaw to a specific malformed instruction, and from there guess as to what the correct data should be.

What method of cleaning did you use?

[fri0701]

Wow! Do you know the story of where these disks came from?

It's a busy time for me, but I'll have to give these a shot when I can.

For the files which were damaged, like the shell files, I wonder if more analysis is possible. Could you run a shell file through a 68000 disassembler and identify which portion of the file came from the damaged sector? If we're very lucky, it may be possible to pinpoint the flaw to a specific malformed instruction, and from there guess as to what the correct data should be.

What method of cleaning did you use?

I bought them from a fellow collector who said that they came with a Lisa 2 that he bought years ago.

The disassembly idea sounds interesting... I'll have to give that a try! I'll post here if I discover anything else.

I'd read on the forums that using alcohol was frowned upon by some, so I started off just using water on a Q-tip going very gently over the disk surface before wiping off with a microfiber cloth. This seemed to make some improvement in the "haziness" of the disk surface, but not in removing any of the suspected-mold spots. After taking several reads and seeing that some of the sectors that were consistently bad, I did try using a bit of alcohol using the same method to see if any improvement could be made, but no luck. I consider myself lucky that *anything* was still readable on these disks, and that we got to them when we did! The magnetic surface really looks terrible, with blotches and streaks everywhere - and they look integrated into the media itself rather than sitting on the surface, if that makes sense.

[fri0701]

After taking a look at the damaged sectors on disk 1, it does look like they're still incorrect despite two reads producing the same data  :-\. For each sector on disk 1, the surrounding code has noticeably fewer differences compared to LOS 1.0 and LOS 1.2, but the sector itself is wildly different. I can only assume this is also the case for disk 2.

I'm admittedly not very experienced with interpreting large chunks of disassembled code, so perhaps someone who's more comfortable with that could take a look and try to figure out what's happening around these bad sectors. Specifically, there are two in "system.os" on disk 1, and two in "shell.Office System" on disk 2. Their locations in the images are noted in the log files I attached in the original post.

[fri0701]

UPDATE:
I was recently allowed access to an original Lisa system that had a different international version of LOS 1.1 installed on its ProFile. I was able to successfully back it up using BLU, and to my relief, realized that it seems to contain the files that have broken sectors on this LOS 1.1B disk set.

I've taken the corresponding 4 sectors and patched my disk images with that data. I don't have a Lisa 1 to test these on - if anyone here has one and wants to give it a go, I'd love to see if my repair attempt worked or not. Please send some screenshots!