> The ATT 6300 was a nice box for the time. The color monitor was really
> something (if you had color). It had graphics capability somewhere
> between CGA and EGA for color and something close to a Hercules card for
> mono.
>
> There is no setup disk or rom bios settings for this machine. Mostly that
> came about with 286-class pc's. I've not yet seen an XT/PC class box with
> chip-stored settings.
>
> You can probably format the hard disk by loading DEBUG from floppy and
> entering the command:
>
> "G=C800:5" - omit the quotes.
It depends -- the DTC cards may have been G=c800:ccc
The AT&T had the hard disk parameters and disk controller stuff in the
main bios rom. The 1.43 could support the AT&T standard DTC controller
and the WD1002 and the later RLL varient.
>
> This command calls and executes a program on the hard disk controller
> which will low-level the drive. Make sure you know the geometry of the
> drive before executing a low-level format as it will prompt you for them.
> The controller will store your geometry on the first track of the hard
> disk and pull from there just prior to booting from it.
The AT&T only supported certain geometries in their bios (they disabled
the hard disk controller's bios rom in most machines).
Some of us who retrofitted the hard disks had to upgrade the ROM's to
the 1.43.
>
> I think the machine's isa buss is 8 bit. You can graft an 8-bit ide
> controller onto it if you like. I don't know a thing about any other
> (proprietary) slots but it sounds like ATT to do something like that.
>
The Olivetti/AT&T/Xerox machine had 16 bit and 8 bit slots. The 16 bit
are Pre-AT. They also handled the byte order differently so there were
other issues.
With a V30 CPU replacing the 8086 they smoked all the XT clones for
performance. They also could be retrofit for EGA or VGA graphics
if you had the right modifications.
They also had an on-board clock that was fed by a NICAD and AT&T's OS
had support for reading the clock at boot time like an AT. They did the
code for this in Juniper Plaza in Freehold, NJ five blocks from
my old house.
However, they did a write back to it at boot time which cost the
content of the seconds register.
In a day of serious debugging I lost 30 minutes due to reboots.
They also forgot much of what they did -- support told me the
motherboard needed to be swapped and that AT&T didn't do the bios code.
This caused AT&T to replace the motherboard. I went through the code with
debug and found where the code was done and changed the write to a
second read (to not change the number of bytes)... This eliminated the
problem. I told this to AT&T hw support -- which was swamped with
motherboard requests after they rolled out the v3.2 dos (which supported
disks with partitions over 20 mb -- in a non-standard pre-dos-3.3 way.)
They were very happy to hear about the clock.
I don't know if there was any official fix.
--Bill
--
bpechter_at_monmouth.com | FreeBSD since 1.0.2, Linux since 0.99.10
| Unix Sys Admin since Sys V/BSD 4.2
| Windows System Administration: "Magical Misery Tour
Received on Tue Oct 17 2000 - 12:50:25 BST