Ham sandwich navigation and radar failure

Roy Smith wrote:
Dan Truesdell wrote:

Someone once told me that the term "winchester" came from IBM, where
their non-removable drive was the 3030. Any truth to that?


That's certainly the story I always heard. Goes back to the IBM-1130
days. I played with an 1130 some, but they were pretty much gone by the
time I came around.

The IBM 3330 (not 3030) was the Winchester and it was introduced shortly after
the System/360 gave way to the System/370. I tested the OS MFT/MVT software for
them back in the early '70s. IIRC it was the first IBM drive that used "rotation
position sensing", so reads/writes could be ordered to minimize rotational delay.

The 1130 had a 2311 drive, I think. Saw 1130s that were used for chip design but
never used one personally. They drove a huge plotter that produced an image that
was photographically reduced to make the masks for the photlithography in chip
manufaturing. The 1130's sister, the 1800, was used as a process controller for
pulling silicon crystals.

Yup, I worked in the old Components Division where all the cancer clusters are
generating lawsuits.

--

Remove SHIRT to reply directly.

"Tarver Engineering" wrote in
:


Somewhere in the closet I've got some CDC-6600 memory. Probably more
interesting from a history of technology point of view, but not as
pretty, so it lives in the closet instead of on the wall :-)

I have an entire 6600 stack.

Young pups!!! The CDC-6600 was what we got to replace the CDC-1604, which
was an upgrade from the CDC-8231, which was to replace the IBM-1620, which
meant we didn't have to use the LGP-30 (with one word of internal memory)
anymore. Eventually we were able to get rid of the Pace 220 (anyone need a
LOT of dual-triodes?).

[Yes, I still remember many long nights trying to figure out where the 6600
performance went, when you were pretty sure it was just some mod to a piece
of PP code, and someone forgot to do a piffer (PFR) in a loop. Gack! The
useless information I still have cluttering up my aging mind.]

-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721
-----------------------------------------------

"Jaap Berkhout" wrote in
:

I am quite sure it had 256 diodes (I modified one to boot not only
from 8" floppy, but from papertape and disk as well). This was on a
PDP 1140.

I know the PDP-8 had the diode boot card. Did the 11/40? For some reason
I thought it had a real ROM. [May have been an option for the diode card -
allowed you to change it. Better than the earlier version where you
actually pried pieces of copper off the board.]

We still have some PDP-11's in operation. Support for a couple of Air
Force contracts. One is running the development software. Two more are
running the test station. But at LEAST we were able, a couple of years
ago, to get rid of the paper tape punch!!! Through a party when that went!

-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721
-----------------------------------------------

"James M. Knox" wrote in message
...
"Tarver Engineering" wrote in
:


Somewhere in the closet I've got some CDC-6600 memory. Probably more
interesting from a history of technology point of view, but not as
pretty, so it lives in the closet instead of on the wall :-)

I have an entire 6600 stack.

Young pups!!! The CDC-6600 was what we got to replace the CDC-1604,
which
was an upgrade from the CDC-8231, which was to replace the IBM-1620, which
meant we didn't have to use the LGP-30 (with one word of internal memory)
anymore. Eventually we were able to get rid of the Pace 220 (anyone need
a
LOT of dual-triodes?).

I wrote my senior project on a CDC-4800, I don't think you can beat that.
A real "main frame" computer.

[Yes, I still remember many long nights trying to figure out where the
6600
performance went, when you were pretty sure it was just some mod to a
piece
of PP code, and someone forgot to do a piffer (PFR) in a loop. Gack!
The
useless information I still have cluttering up my aging mind.]

I drew the install for RPL's Cyber 180 that RPL replaced their 6600 with.
Hadly any of the stacks survived, because people wanted the individual
memory cards. (12 ea)

"James M. Knox" wrote in message ...
"Jaap Berkhout" wrote in
:

I am quite sure it had 256 diodes (I modified one to boot not only
from 8" floppy, but from papertape and disk as well). This was on a
PDP 1140.

I know the PDP-8 had the diode boot card. Did the 11/40?

Yep.

For some reason I thought it had a real ROM. [May have been an option for the diode card -
allowed you to change it. Better than the earlier version where you
actually pried pieces of copper off the board.]

The diode card is a real rom. The older 11's all booted out of UNIBUS memory.
I'm sure someone made a UNIBUS card with a ROM chip on it, but I never saw one.
It only takes a few instructions to load sector zero off the disks like the RK05 and jump
to it. From there you have 512 bytes to play with to retrieve the rest of the boot image
off the disk.

We still have some PDP-11's in operation. Support for a couple of Air
Force contracts. One is running the development software.

Crikes, even back when I was running around the Army recycling people's no longer
suitable for real work PDP-11's into internet routers, I was cross compiling from a
VAX. Of course VAX's are pretty much gone these days as well.

Now the key is, do you know what the MARK instruction does.

Roy Smith wrote:
Dan Truesdell wrote:

Someone once told me that the term "winchester" came from IBM, where
their non-removable drive was the 3030. Any truth to that?


That's certainly the story I always heard. Goes back to the IBM-1130
days. I played with an 1130 some, but they were pretty much gone by the
time I came around.

I remember the "Winchester" as being a single platter sealed drive
that gave you 30 meg with 30 ms access time. The drives quickly
evolved into faster/higher capacity/smaller size units with the
name Winchester continuing to be associated with the "sealed"
method of construction. First actual 30-30 I remember seeing was
an option on a line of Ohio Scientific microcomputers.

Tom Pappano, PP-ASEL-IA, Ohio Scientific C1P 32k 6502 Dual floppy/ASR33

"Tarver Engineering" wrote in
:

I wrote my senior project on a CDC-4800, I don't think you can beat
that. A real "main frame" computer.

Actually, that's way newer than most of the ones I worked on. Heck, it had
RTL IC's and (if I recall correctly) those wierd back-to-back cards. Not a
vacuum tube in sight.

-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721
-----------------------------------------------

"Ron Natalie" wrote in
:

The diode card is a real rom.

Or could be called an FPROM, since we changed then in the field.

Now the key is, do you know what the MARK instruction does.

I know what it did on the IBM-1620. Set the mark bit at a digit
address. [For those "newbies" lurking - the 1620 didn't have words. It
had digits. Banks of 20,000 each (we could only afford one bank). Each
digit took six bits. Four were for the decimal digit (no HEX spoken
here!). One was a mark bit, and one was ... or durn, "stop" bit?. Each
instruction was terminated by setting one or both of these bits.

You could clear memory by doing a write of an instruction to the next
address, that overwrote the stop/mark bits with an instruction. You
could clear a bank of memory in about 1/8th second!!! Amazingly fast.

To give you an idea of just how fast this machine was, I was pretty good
at debugging through my program with the run/step switches. You could
look down at your program, go "Hmmm, I know it's probably 20 or 30 lines
to the next section I need to step through." and do a quick RUN, then
press STOP. Time it right and you could pretty much estimate how many
lines of code you had executed. [Try THAT with your Pentium 4!]

I don't remember a mark instruction on a VAX, or even a PDP-11.

Of course, there are always the instructions like:

SPI -- Shred Programmer Immediate

or

HCF -- Halt and Catch Fire


-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721
-----------------------------------------------

"James M. Knox" wrote in message ...


I don't remember a mark instruction on a VAX, or even a PDP-11.

Not all the PDP-11's had it. It was kind of a useless thing. It was an oddball
subroutine linkage instruction that you put on the you'd jump to it and it swapped
a register and jumped again.

"James M. Knox" wrote:
I don't remember a mark instruction on a VAX, or even a PDP-11.

I don't remember the details, but it did something funky with the stack
pointer and R5 which only worked right if you weren't running separate
I/D space. Had something to do with jumping back and forth between
co-routines (early hardware support for threading?).