Disk archival

[Jim Leonard]

Tony Duell wrote:
> I would strongly discourage the use of Teledisk for this. The file format 
> is proprietary, and although there have been some attempts to 
> reverse-engineer it, AFAIK the full details aren't know (particularly of 
> Teledisk compressed files). 

F-prot has been able to scan inside teledisk images for over a decade.  One of 
the authors graciously let me look at the source a while ago and the 
compression is nothing more than LHA (ie the free lharc source by yoshi).  I 
might still have the source somewhere, but it's buried in email archives going 
back a decade.  If someone needs it to do something useful, let me know 
off-list and I can try to dig it up.  (It would give me an excuse to finally 
organize and index my email archives!)

> If you use Teledisk, you _have_ to use an 
> MS-DOS PC to re-write the data to a physical disk (and a PC with the 
> appropriate type of floppy drive),  you _have_ to get Teledisk (last time 
> I checked, it wasn't free, it was shareware, but you could no longer 
> register it...). etc. If your PC can't write FM formats, then you're 
> stuck (even if you've got 100 machines that _can_ write such formats, and 
> which can get data from a PC disk or similar). 

The problem with various methods of archiving diskettes is that they are all 
proprietary methods with drawbacks:

Disk2FDI: works with any disk, requires dual-drive PC or special cable, file 
format is bitstream and extremely complicated to work with
Central Point Option Board: works with FM/MFM/GCR disks, requires special hardware
Catweasel: works with any disk, requires special hardware, no mature software 
exists specifically for archiving, no mature/stable file format
Teledisk: works with MFM disks only, file format not published, limited
CopyIIPC+Snatchit: works with MFM disks only, requires old PC, file format not 
understood

Nobody has tackled this to any great capacity.  Disk2FDI is the best option 
right now since the file format is well described, but at the same time the 
files produced by Disk2FDI are nothing more than bitstreams that are nearly 
impossible to work with without knowledge of the controller hardware.  (In 
fact, the only application other than Disk2FDI that works with FDI images is 
WinUAE, since it emulates the disk controller at all levels.)

When I started archiving my 5.25" software in the 1990s, much of it 
copy-protected, I settled on the Central Point Option Board.  This was for 
several reasons:  I knew I would always have 8088-80386 hardware to use with 
the board; I only had MFM/FM/GCR disks to archive (ie no 
hard-sectored/proprietary/goofy formats); the file format is a hybrid of raw 
(gap length, header info) and cooked (actual sector DATA stored as bytes) which 
was the most useful to me (without any knowledge of the archive file, I can do 
"strings filename.img" and read plaintext data.  Whenever I saw an option board 
for sale @ $20 or less (usually not on ebay), I would pick it up as a spare.  I 
have 5 "spares" now :-)

> Any such archive must be made as accessible as possible, and that means 
> using documented file formats. To be honest, I'd rather have to write my 
> own software than do battle with a proprietary PC program.

But then *your* program would be proprietary :-)

What is needed, and has been attempted and mostly failed by several different 
people, is this:

- A standard method of describing not only what is stored on a diskette, but 
HOW and WHY (to make sense of all the gap/header/sync/etc. bits)
- A common way of getting that information off the disk, which will probably 
require interfaces to PCs or MACs because the original hardware may not be able 
to provide every single flux off the disk
- Well written "disk controller emulation" utility programs to translate the 
archived stream of bits into useful data, the highest priority being the 
ability to extract individual files

As you can imagine, everyone has their own idea on how to do this.  Not 
surprisingly, the people who have had the most headway with projects like this 
have been software pirates (reformed, grown-up, 9-to-5 job software pirates) 
who -- though they mean well -- have a very limited vision of what archival is. 
(Meaning, if you can read the files, and run the program, it's good to them. 
Who knows what kind of alterations had to be made to get to that point, but 
they don't care.  I personally WANT copy-protection to remain intact, etc.  I 
want exact copies.)

Until some common standard is defined and widely accepted and applies equally 
to every type of diskette-based magetic medium, I'll just keep on doing what 
I'm doing that works for me, and I'll contribute my work to the global standard 
if/when it exists.  Just because there isn't a standard doesn't mean I should 
hold off archiving.
-- 
Jim Leonard (trixter at oldskool.org)                    http://www.oldskool.org/
Want to help an ambitious games project?             http://www.mobygames.com/
Or check out some trippy MindCandy at             http://www.mindcandydvd.com/