On Wed, 2011-11-02 at 23:32 +0000, Kieron Wilkinson wrote:
People, until
recently, couldn't create IPF files themselves. The IPF library source
is now available, so people now can do it if they want, so the comment
doesn't really apply anyway.
Correct me if I'm wrong, but I was under the impression that:
1) The IPF library was read-only
2) There were no publicly-available documents covering the IPF file
format
3) No documents have been released as part of the source dump (thus
satisfying (2))
Thus, creating an IPF would involve reverse-engineering the format from
the source code (which I have not examined closely, but appears to be
somewhat Windows-centric and lightly commented based on the few files
I've examined). Once the format was r/e'd, you could then write a
library to create IPFs...
Seems a lot of work. I'd rather put the effort into writing a .MFI (MESS
Floppy Image) R/W library...
I admit that we (SPS) have taken too long to do some
things - stuff now
rectified. But we seem to get bashed in here whenever we or KryoFlux
is mentioned for reasons I don't really understand.
I'm going to pull out a 9mm and shoot myself in the foot here.
* Kryoflux and DiscFerret are two separate things.
* Kryoflux does a good job of imaging most floppy discs
* Key differences between DiscFerret and Kryoflux:
* DiscFerret uses "store and forward" acquisition. KryoFlux (AIUI)
streams data directly to the PC. This means DiscFerret is slightly
slower to read and write discs (by around 1 minute).
* DiscFerret has more usable I/O lines, plus the 4-bit bidirectional
High Speed I/O link (four pins which may be programmed for any purpose;
GPIO, SPI, etc. and used for I/O expansion)
* DiscFerret captures data at a much higher resolution (approx. 4x
Kryoflux's peak rate, and twice the Mk4 Catweasel). This means you can
image hard drives (ST506 type) and 2.88MB floppy discs, with suitable
adapter cables.
* DiscFerret has full support for both hard-sectored and soft-sectored
media, including reading individual sectors. CW could do this too, AIUI.
On the down side, DiscFerret is currently lacking in software. I have a
tweaked version of CWTOOL which sort-of works with DFI images, but falls
over quite a lot. Disc reading works, but writing needs work (mainly
timing characterisation and modelling).
the KryoFlux personal edition is
in fact sold at a massive loss when you factor in the time we spend on
it.
The DiscFerret is no better. They're sold more or less at cost...
--
Phil.
philpem at philpem.me.uk
http://www.philpem.me.uk/