FrameMaker 3 files

[subscribe at cuff.ca]

Thanks to everyone for their help. You've all been generous with it!

 =-=
to: Jeremy H. Griffith

Thanks for the tip about file length. As it happens, all five files have sizes evenly divisible by 1024 bytes. I also examined a couple of files to see if they ended (or even began) with a spurious "0D 0A" (or either, singly), and that's not the case either. Good thought, though. I'm going to add this tip to my toolbox.

 =-=
to: Frank Stearns

I'm betting you're right; the files are corrupt. At least I have been able to resurrect some of the information, even if I haven't been able to fully recover the files.

I asked the developer who supplied the files how the files were stored, and he told me that the files were stored on hard drives (no further information; see next paragraph), in a CVS repository. There's more--the CVS repository was imported from an older revision history system. The history for the older system notes that the files were added to the system in 1993, but doesn't have information on modifications. (Which could mean either no modifications or no tracking.) The files were checked in to the older system by someone other than the original authors. The files have not been modified since they were added to CVS.

Given all that, our developer notes that it is perfectly possible that the files were corrupted before they were introduced to the older system, with the usual caveat about CVS's habit of messing with CRs and LFs. (I did try altering the files to compensate for this, with no result.) One additional complicating factor is that several of our file servers used to be on-site, and that quite a lot of them are now networked drives in distant cities.

Sorry if this isn't quite the information you were hoping for.

 =-=
to: Shlomo Perets

Thanks for the FM3 file. It opened easily enough without my having even to resort to "heroic open." As far as I'm concerned, this is the final nail in the coffin--the files are corrupted.

The good news is that the preliminary work I put into resurrecting data from the files hasn't been wasted. In the end, only the diagrams have been lost, and some of the context of the largest file is a little tricky to reconstruct. That's not too hard to bear at all.

 =-=
to: all those who wrote me off-list

Thanks! I will reply soon if I haven't already. (I'm temporarily separated from my mail client, and am relying on webmail.)

Doug