11 Oct
2008
11 Oct
'08
11:28 p.m.
Does it fail to see any sectors at all?
Correct. I haven't tried a catweasel on it yet to see
what the details of the sectors are.
12 Oct
12 Oct
7:50 a.m.
Does it fail
to see any sectors at all?
Correct. I haven't tried a catweasel on it yet to see
what the details of the sectors are. 
10:25 p.m.
Dave Dunfield wrote:
most likely there is some other factor in the 8-sector
format
which makes it incompatible with the PC disk controller...
I don't remember the particulars of the 3b1 disk formats, but they used
a 2797 FDC, which is certainly capable of reading and writing some disk
formats that the 765 FDCs can't deal with. As you say, there must be
more different than just the 8 sectors per track, as IBM DOS 1.0 used 8
sectors per track on the PC.
Possibly they are using a non-standard code for the sector length in the
address fields.
Eric
13 Oct
13 Oct
2:01 p.m.
-------------- Original message from Eric Smith <eric at brouhaha.com>:
--------------
Dave Dunfield wrote:
When I image disked my set, it found both 8 and 10 sector disks. The 3B1 can
format in either 8 or 10 sectors. It looks like the main software is all in 10 sector
disks and the other add on software is in 8 sectors. The 2 disks that failed to image
where from the same add on package, don't remember which and they showed
up as 9 sectors ??? Anadisk finds the same and can read the first 8 sectors and
errors on the 9th.
I might try tonight to reload the OS on another drive to see what the Images them
selves produce.
Jerry
Jerry Wright
g-wright at att.net
most likely there is some other factor in the
8-sector format
which makes it incompatible with the PC disk controller...
I don't remember the particulars of the 3b1 disk formats, but they used
a 2797 FDC, which is certainly capable of reading and writing some disk
formats that the 765 FDCs can't deal with. As you say, there must be
more different than just the 8 sectors per track, as IBM DOS 1.0 used 8
sectors per track on the PC.
Possibly they are using a non-standard code for the sector length in the
address fields.
Eric
2:15 p.m.
When I image disked my set, it found both 8 and 10
sector disks. The 3B1 can
format in either 8 or 10 sectors. It looks like the main
software is all in 10
sector
disks and the other add on software is
in 8 sectors. The 2 disks that failed to
image
where from the same add on package, don't remember which and they showed
up as 9 sectors ??? Anadisk finds the same and can read the first 8 sectors and
errors on the 9th.
As I recall, earlier distributions of the system were all
on 8 sectored disks. But as the distributions grew, they
started using 10 sectors to fit more on the disk. But I
do remember that at least one of the disks was always 8
sectors. I'm inclined to say that might have had to do
with the on-board boot code. The PROM code may not have
been able to boot from a 10 sectored disk, but that's as
much or more speculation as it is memory.
BLS
4:41 p.m.
On 13 Oct 2008 at 18:01, g-wright at att.net wrote:
When I image disked my set, it found both 8 and 10
sector disks. The 3B1 can
format in either 8 or 10 sectors. It looks like the main software is all in 10 sector
disks and the other add on software is in 8 sectors. The 2 disks that failed to image
where from the same add on package, don't remember which and they showed
up as 9 sectors ??? Anadisk finds the same and can read the first 8 sectors and
errors on the 9th.
Depends on what the 9th sector display as. There's a Windoze-related
glitch in the Anadisk sector map code to detect "orphan" sectors;
i.e., those without an address header. If the sector shows up in the
map as "109" or some such, you're seeing the artifact. Try it again
with honest-to-gosh real-mode MS-DOS.
Believe it or not, the "orphan" sector was someone's idea of copy
protection back in the day. Shows up on a "Read Track" operation,
but is otherwise inaccessible. Similarly, some vendors used to put
about 50 orphan IDAMs on a floppy with no data sectors. Or put 2
sectors with identical IDAMs on a track--read one and then read the
other immediately and compare the results. As well as sticking stuff
in the inter-sector gap to identify the disk (I think Harvard
Graphics did that one).
Ah, the good old days.... :(
Cheers,
Chuck
14 Oct
14 Oct
2:22 p.m.
New subject: Imaging, AT&T 3B1/UNIXPC/PC7300 software
-------------- Original message from "Chuck Guzis" <cclist at
sydex.com>: --------------
On 13 Oct 2008 at 18:01, g-wright at att.net wrote:
snip >>>>>>>>>>>>>
> errors on the 9th.
Depends on what the 9th sector display as. There's a Windoze-related
glitch in the Anadisk sector map code to detect "orphan" sectors;
i.e., those without an address header. If the sector shows up in the
map as "109" or some such, you're seeing the artifact. Try it again
with honest-to-gosh real-mode MS-DOS.
Believe it or not, the "orphan" sector was someone's idea of copy
protection back in the day. Shows up on a "Read Track" operation,
but is otherwise inaccessible. Similarly, some vendors used to put
about 50 orphan IDAMs on a floppy with no data sectors. Or put 2
sectors with identical IDAMs on a track--read one and then read the
other immediately and compare the results. As well as sticking stuff
in the inter-sector gap to identify the disk (I think Harvard
Graphics did that one).
Ah, the good old days.... :(
Cheers,
Chuck
Well Chuck you had me thinking for a while but copy protection on
the print drives disks did not seem right. I was using real DOS with
image disk and Anadisk.
So I decided to try to use the Images had made. So I found a different
HD drive, booted the OS from floppy and it all loaded just fine.
The 2 add-on disks that would not image, I made copies on the
UNIX PC and they loaded just fine. I then tried these copies with Image
disk and that worked. So I would guess it may be a case of age and/or how
they where made. All of the foundation disks are 10 sectors and the add-on
and boot disks are 8 sectors. All of the Images where read and written
using a 360k floppy drive and DSDD media.
- Jerry
Jerry Wright
g-wright at att.net
13 Oct
13 Oct
4:51 p.m.
On Mon, 13 Oct 2008 g-wright at att.net wrote:
where from the same add on package, don't remember
which and they showed
up as 9 sectors ??? Anadisk finds the same and can read the first 8 sectors and
errors on the 9th.
Some systems that use multiple different numbers of sectoras on a track
will format 9 (or 10), and then depending on which format is desired, only
use 8 (or 9).
Also, if you reuse disks that were previously formatted for 9 sectors per
track (such as PC-DOS "360K"), some systems will see that the 8 sectors
that they need are present, and mat choose to not reformat the disk.
1:42 p.m.
On 12 Oct 2008 at 6:50, Dave Dunfield wrote:
Then most likely there is some other factor in the 8-sector format
which makes it incompatible with the PC disk controller...
Given (a) that these are 48 tpi 5.25" DD diskettes and (b) that the
3B1 uses a WD2797 at standard data rates, you should *at least* see
returns from a Read ID operation. ISTR that while the 2797 could
format using a variety of DAMs, there's only one recognized MFM IDAM
and it must be preceded by three A1s to be recognized reliably.
If you're not seeing at least the address headers, it doesn't look
good.
Cheers,
Chuck
Does it fail to see any sectors at all?
Correct. I haven't tried a catweasel on it yet to see
what the details of the sectors are. 
4:52 p.m.
On Mon, 13 Oct 2008, Chuck Guzis wrote:
On 12 Oct 2008 at 6:50, Dave Dunfield wrote:
Then most likely there is some other factor in the 8-sector format
which makes it incompatible with the PC disk controller...
Given (a) that these are 48 tpi 5.25" DD diskettes and (b) that the
3B1 uses a WD2797 at standard data rates, you should *at least* see
returns from a Read ID operation. ISTR that while the 2797 could
format using a variety of DAMs, there's only one recognized MFM IDAM
and it must be preceded by three A1s to be recognized reliably.
If you're not seeing at least the address headers, it doesn't look
good.
So, does anyone know if we can count on sometime in the next few months
having a complete set of install disk images that can be written to floppy
with or without a catweasel?
This made me think of a possible alternative. How hard would it be to do
a fresh install on a 3b1, then dump the hard drive? Would that eliminate
the need to deal with weird formats? Would that be acceptable for the
different variations of machines?
Note, I have a catweasel specifically for weird formats like this.
--
David Griffith
dgriffi at cs.csubak.edu
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
Does it fail to see any sectors at all?
Correct. I haven't tried a catweasel on it yet to see
what the details of the sectors are. 
14 Oct
14 Oct
8:15 a.m.
On Mon, 13 Oct 2008, Chuck Guzis wrote:
On 12 Oct 2008 at 6:50, Dave Dunfield wrote:
Then most likely there is some other factor in the 8-sector format
which makes it incompatible with the PC disk controller...
Given (a) that these are 48 tpi 5.25" DD diskettes and (b) that the
3B1 uses a WD2797 at standard data rates, you should *at least* see
returns from a Read ID operation. ISTR that while the 2797 could
format using a variety of DAMs, there's only one recognized MFM IDAM
and it must be preceded by three A1s to be recognized reliably.
I had no problems at all reading that diskette on my Linux box and
capturing a 'dd' image.
Al, I sent this to you but never received any response. Did you not
receive it?
Steve
--
Does it fail to see any sectors at all?
Correct. I haven't tried a catweasel on it yet to see
what the details of the sectors are.
5761
days inactive
5763
days old
10 comments
9 participants
participants (9)
-
aek＠bitsavers.org -
blstuart＠bellsouth.net -
cclist＠sydex.com -
cisin＠xenosoft.com -
dave06a＠dunfield.com -
dgriffi＠cs.csubak.edu -
eric＠brouhaha.com -
g-wright＠att.net -
snhirsch＠gmail.com