Re: GS/OS FST files?

["RedskullDC" <RDC@oz.org>]

Hi schmidtd,

"schmidtd" <schmidtd@my-deja.com> wrote in message 
1176819049.856159.200630@e65g2000hsc.googlegroups.com">news:1176819049.856159.200630@e65g2000hsc.googlegroups.com...
> On Apr 17, 8:46 am, "RedskullDC" <R...@oz.org> wrote:
>> ProDOS-8 I am not sure about. I don't have much in the way MLI doco. Is
>> anything other than the BLOCK interface exposed?
> Worth and Lechner to the rescue once again.
> http://www.apple-iigs.info/doc/fichiers/beneathprodos.pdf
>

I have skimmed through that book before, and also the prodos tech manual 
from Apple.

It would be possible I guess to patch the MLI start vector, examine the 
parameter lists to see if the request was for our device, and re-route to 
the approriate routine (preferably in a ROM on the card).

Most of the MLI calls would be redundant, and could just return CLC to 
indicate success. (BLOCK related).

Random file access wouldn't be possible with the firmware on the USB host 
chip (as it stands), so all the File position calls would be useless.
They would have to return an error. Ex: Get/Set Mark, EOF etc.

That would leave:
Create, Destroy, Rename, Set/Get Prefix, Set/Get File Info, Online, 
Open/Close, Read/Write , Set/Get Buffer.

Not too much of a drama....

Problems are:

I want as much existing software to work as possible, otherwise it is a pain 
in the.....

Firmware on the interface card can't identify itself as a "block" type 
device, since it isn't.
It could identify as a $38,$18,$01,$7x (Mass Storage device), but I would 
imagine the Basic Interpreter makes assumptions about this being a "block" 
device.

If the firmware identifies as a "non-block" device......
Does the BI keep internal tables regarding which MLI calls are valid for the 
different kinds of devices?
I imagine it wouldn't try to call the firmware on an 80-col card to get a 
directory listing for example.

I don't really want to hack around inside the BI if possible.

Red