Re: Help me get my childhood Apple //e running again

[Tom Moertel <tmoertel@gmail.com>]

On Monday, April 30, 2012 9:43:37 AM UTC-4, Linards Ticmanis wrote:
> Well, DOS doesn't use the upper RAM "under" the ROM while ProDOS does.
> [...]
> 
> IIRC, ProDOS does the GCR encoding/decoding of most of the sector
> in-place but uses an internal buffer for the last part. But I'm not
> certain if that "last part" was the high (or low?) bits of every byte or
> if it was the last bytes of the sector. Still that could explain your
> problem. If the GCR decoding happens incorrectly, of course the CRC
> would be calculated over the wrong bytes and would match those wrong
> bytes, so the ADTPro server on the host side would never know anything
> went wrong...

Linards, thanks for this information!

My best hypothesis so far was that something was going wrong with 6+2 decoding, since only the lower two bits were ever affected, and that, since ADT 2.4 uses DOS and ADTPro ProDOS, the way ProDOS decodes disk data differently from DOS might be a link in the chain that leads to the problem I'm seeing.  But I couldn't find the other link in the chain until you said that ProDOS uses higher memory for part of the decoding.  Now all the evidence fits the hypothesis: If some higher RAM that DOS ignores were bad, we would expect to see everything that I'm seeing.

I've actually got a 512K RAMWORKS card in the AUX slot, so maybe I'll try swapping it out with an Apple 64K card if I can find one.

Thanks for your help!

Cheers,
Tom