Apple again: DOS 3.2 image should handle header-only sectors

[TeaRex]

The disk formatting code of Apple DOS 3.1 / 3.2 does something fairly odd: It writes only the sector headers, with no sector bodies. Any particular sector's body is created the first time that particular sector has data written to it. This means that on such disks it is normal for unused sectors to be "bodyless", but kryoflux currently reports those sectors as non-existing with "-i7", and if any track is completely unused, it is reported as "<unformatted>".

The created sector image is fine as it will have zero bytes in those places, but the messages are firstly confusing and secondly not as helpful as they should be (e.g. you can't tell whether a track is really unformatted or whether it just doesn't have any already-used sectors on it). I suggest treating those "bodyless" sectors in -i7 format as existing, and containing all zeros.

Note that this does NOT affect Apple DOS 3.3 / Apple ProDOS format(-i8, -i8a), nor the 3.5" ProDOS/Macintosh formats (-i9).

[IFW]

Thanks.
As usual, it would be very useful to post stream dumps of such tracks, e.g. partially used, and completely unused.

I can see one problem however: there would be no way of seeing genuine errors resulting in missing sector data vs unused blocks...

[TeaRex]

Just screwed up my drive, so I'll post streams after I fix it. OK?

[IFW]

Thanks and good luck fixing it

[TeaRex]

dos32.7z

Stream Data

(240.25 KiB) Downloaded 219 times

OK here are the first 4 tracks of a real freshly formatted Apple DOS 3.2 disk. The first two should be full; the third, partially used (9 of 13 sectors); the fourth, empty. Dumped with an Epson SD-600 drive for now as my main "flippy capable" drive is still broken. I hope you can handle .7z?

[mr.vince]

I am sure we can.

[TeaRex]

Ok, I'll make sure to use something more obscure next time... paq8px_v66 maybe?

[mr.vince]

Don't you dare, young padawan!