[Coco] Floppy - less boot , was Re: VI and VIRQ

[Gene Heskett]

On Wednesday 14 May 2008, Mark Marlette wrote:
>Gene,
>
>80T=96TPI
>40T=48TPI

For 5.25" drives only Mark, all 3.5" drives, even the original single sided 
ones, are 135 tpi.

>I always thought to get a 48TPI disk out of a 96TPI drive that you
>would have to double step. Thus the driver would have to know what the
>drive is.

That is a read-only situation mark, because the heads in the 5.25" 80 track 
drive are narrower, they do not erase the full track width when doing a 
write, so at that point, the real 40 track drive with its wider heads, will 
be reading both the narrow 96 tpi track, and the remains of the wider 48 tpi 
track, with predictable gigo results.  And several times, I've not been able 
to re-read what was just written, even with the narrow head of the 80 track 
5.25" drive.

To format and use an 80 track 3.5 drive, do NOT change the dns value in the 
descriptor, only the cyl(inders) to 50(=80 decimal) tracks.  There is never 
any need to double step the 3.5" drive, they are all 135 tpi.

If the descriptor in memory has a higher dns value than is read from the disc, 
then the driver will double step the drive assuming it has found a 40 track 
disc in an 80 track drive.

This only applies to 5.25" 80 track drives, leave the dns set for whatever 
=mfm (2?, memory, second thing to go you know) for ALL others.

To me, since that condition does not also set that disc to read-only, changing 
the dns value is at best a mixed blessing that has lots of barbs attached. 
You forget, and try to write to that disc and boom, disc format is wrecked, 
and it must be reformatted in the right drive to recover.  Your data is of 
course history in that event.

I almost wish that particular workaround had never been coded.

OTOH, its handier than sliced bread and bottled beer when you need it.

>That is part of the reason why it is hard to make a CoCo disk on a PC
>without a 5.25" 48tpi drive.

Which has prompted me, due to 3.5" disc interchange sloppiness, to carry a 
3.5" drive back and forth as needed.  I have 3, any one will do, but they 
can't read each others discs.  Its a PITA, but it works great. :)
But fdutils-5.5 seems to require that I format the disk to be used in that 
manner on the coco. fdutils-5.4 works fine ISTR.  Linux box here of course.
>??
>
>I don't mess that much with Emulators/floppy drives. I use the real
>thing, still have over a hundred cocos in the house. :)

I don't emulate either. :)  Except for printing, I just bought a Brother 
HL-2140 B&W laser for the coco.  Its a GDI printer of course, so /p feeds a 
serial-usb convertor, which feeds a 16 foot usb extension cable/hub up to 
here.  The printer is sitting on the coco's desk with another such 16 foot 
usb extension cable/hub driving it.  Brother was very nice in telling me 
where to find the linux drivers for it, and they work.  Now, if the coco 
could just output 20 pages a minute. :)

The next problem there is in constructing a pipeline from /dev/ttyUSB1, where 
I can capture the coco's output just fine, and feed that back 
out /dev/usblp1, with a2ps and gs doing the translations from the coco's POA 
(Plain Old Ascii) to the gfx image the printer expects.  Its fun, depending 
on ones personal definition of 'fun' of course.  I may have to write a script 
of sorts to paginate it into page sized pieces & feed it back out as 
individual pages to lpr.  Or bring my C skills up to date, they are pretty 
rusty.

>Mark
>
>Quoting Gene Heskett <gene.heskett at verizon.net>:
>> On Wednesday 14 May 2008, Mark Marlette wrote:
>>> Diego,
>>>
>>> I don't think using an 80t will work unless the boot track module is
>>> also for that drive. The step sizes are different, 80 is 1/2 40's.
>>>
>>> I don't think there is an 80 track boot track. I would think that it
>>> could be made pretty easy. Hardware guys will always say that. :)
>>
>> No reason why not Mark.  Unless THAT particular 80 track descriptor
>> also has a
>> different value in its dns setting.  If an 80 track format is made by
>> dmode /d? cyl=50, to the 40 track DS descriptor, than this is correct for
>> pure 80 track drives of either size, 5.25 or 3.5.
>>
>> The only reason for changing the dns setting to set the next bit up, is so
>> that if a 40 tack disc is inserted in an 80 track 5.25" drive, this diff
>> will be found in LSN0, and the driver will then double step the drive to
>> read that 40 track disc correctly.  Again, that is the only reason.
>>
>> Unforch, I found it is hard to remember that such a disc cannot be written
>> in that 80 track drive without wrecking the disc, and have wrecked several
>> that way.  The driver really should, when it finds itself in the
>> situation, do whatever it takes to duplicate finding a write protection
>> tab on the disc, and refuse to do the write, just to slap big dummies like
>> me upside the head with the error message.
>>
>> There is no difference in the 3.5" disc's track widths as they are all 135
>> tpi, so in the 3.5" case, dns is never touched from the default found in
>> the older descriptors.
>>
>>> Regards,
>>>
>>> Mark
>>>
>>> Quoting Diego Barizo <diegoba at adinet.com.uy>:
>>>> I was using d0_80 because my real drive 0 is an 80t, I guess that
>>>> explains the different behavior when booting from a virtual floppy and
>>>> from a real disk.
>>>> The script is formatting a SS, but I was merging like this:
>>>> merge ../MODULES/BOOTTRACK/rel_80 ../MODULES/BOOTTRACK/boot_ide
>>>> ../MODULES/BOOTTRACK/krn>-bttemp
>>>> Is this one the correct one?
>>>> merge ../MODULES/BOOTTRACK/rel_80 ../MODULES/BOOTTRACK/boot_1773_6ms
>>>> ../MODULES/BOOTTRACK/krn>-bttemp
>>>>
>>>> Damn, it's past midnight... to late to try it now, gonna have to wait
>>>> 'till tomorrow :-(
>>>>
>>>> Diego
>>>>
>>>> Mark Marlette wrote:
>>>>> Diego,
>>>>>
>>>>> Quick look here.
>>>>>
>>>>> Looks like you have the 80T /d0 descriptor selected. Select d0_35s.
>>>>>
>>>>> Then you need to make sure your script is formatting as SS.
>>>>>
>>>>> Then make sure that you merge the correct loader for SS floppy. I
>>>>> have done it, been awhile but it does work.
>>>>>
>>>>> Hopefully that helps.....
>>>>>
>>>>> Mark
>>>>> Cloud-9
>>>>>
>>>>> Quoting Diego Barizo <diegoba at adinet.com.uy>:
>>>>>> Now that it comes to mention. I've been trying to create a custom
>>>>>> 1s35t boot disk for my HDB-DOS / SuperIDE setup, completely
>>>>>> unsuccessfully so far.
[...]


-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
"It might help if we ran the MBA's out of Washington."
-- Admiral Grace Hopper