"Universal Boot Initializer" ?
On Sun, 15 Jan 2012, Vladimir Ivanov wrote:
>
> Steve,
>
> I am not totally sure what you're after, but if I got your questions right:
>
>
> On Sat, 14 Jan 2012, Steve Nickolas wrote:
>
>> Weeeeeeeeell...
>>
>> Of the first 3 tracks, here's usage for a DOS 3.3 48K slave:
>>
>> 0123456789ABCDEF
>> 0##### ##########
>> 1################
>> 2#####
>
> Is T0/S5 missing for a reason?
>
>> Would it be possible to change usage to something more like this?
>>
>> 0123456789ABCDEF
>> 0#####
>> 1################
>> 2##### ##########
>
> Moving part of T0 to T2 is just a matter of loader for booting purposes.
> "INIT" would have to be rewritten, but it's screwed anyway since track #0 has
> to be written with non-standard format.
>
> As we already agreed to, "INIT" is prime sacrifice. ;-)
So it has to be done with a separate program. The tool wouldn't fit in
INIT space anyway.
>> and then stuff two versions of the first 5 sectors onto the same track 0?
>> (i.e., a 16-sector version and a 13-sector version).
>
> Ok, I guess you want to keep T1..T22 in 16-sector format, only T0 is funny.
I dunno if it'd work that way. Might have to be a 4-track 13-sector DOS.
> Here's something even more easier to do and maintain:
>
> Put one 16-sector and one 13-sector S0 on T0. Fill the rest of the track with
> "macro" sectors encoded in 18-sector format, just as we discussed recently.
> The beauty of this approach is that the code in both S0 will be exactly the
> same for 13 and 16-sector cases - it will contain loader for the "macro"
> sectors from T0. In case 256 bytes aren't enough for such a loader (and you
> can't rely on the bootstrap ROM), revert to some simple scheme, even 4&4. But
> the layout of 18-sector "macro" sector is very easy and straightforward.
>
> Again, you write T0 in single pass.
>
>
> Now, the only problem with this scheme is that you want to have 16-sector
> format with 13-sector P5/P6 PROMs on the disk controller. Not reliable and
> you can't expect (I think) that users will have lucky combination of
> 16-sector sequencer PROM and 13-sector bootstrap PROM. (Consult Sather's
> "Understanding the Apple II" for details)
Which is why these disks usually run some form of DOS 3.2 (Bob
Sander-Cederlof's assembler documentation for an early version actually
mentioned using a 16/13 DOS 3.2.1), and one I've seen runs a form of 3.1.
>
> So, to be on the safe side, you have to keep the whole disk in 13-sector
> format, and the "macro" portion of T0 to be in either something analogous to
> 18-sector format but with 5&1 encoding, or plain 4&4 for simplicity.
>
Yeah.
-uso.
|