🚀 go-pugleaf

Guys,
I started programming the IBM RAMAC 305 in 1958.
It had 100 positions of core memory...that was used
as a buffer to read from the disk files to the drum.
Instructions were executed from the drum which
basically held 260 program instructions.  Of course
if you were not a knot head you could set your programs
to read more instructions from disk to drum and execute
many programs, as long as you didn't exceed five million
characters....that was all the disk could hold.  Later
versions of the 305 ramac could have more than one
disk file...last 305 that I wrote programs for had two
disk files storing ten million characters.  There were plug
boards for main processor, printer, card punch and the
card reader.  Most were simple, the main plug board could
be extremely complex.  I remember one progy that had
a 100 way split wire going to "program advance one".

Jim Thomas wrote:

> For Mike:
>
> lwin> From: lwi...@bbs.cpcn.com (lwin)
> lwin> Newsgroups: alt.folklore.computers
> lwin> Subject: Re: IBM 350/1405/1301/1302 disks
> lwin> Date: 26 May 2000 02:08:37 GMT
>
> Jim > (Mike also thinks the disk was called the 305 - the same as the system
> Jim > name.)
>
> lwin> 305 and 350.
>
> lwin> I forget which, but the disk itself had one name, and the system it was
> lwin> a part of it had the other.
>
> lwin> Note that this system, while technically a computer with a programming
> lwin> language, had all of _80_ characters of core memory!  It was mostly
> lwin> "programmed" by wiring plugboards for its various peripheral devices.
>
>         I dimly recall the core memory being 100 characters (we used to
> have a stack from a 305 in a plexiglass box as a "teaching aid"), but
> it was not the main memory anyway. It was the "sector buffer". The main
> memory of a 305 was a drum. If you think for a while about the problem
> of transfering data between a spinning disk and a spinning drum, the need
> for either a sector buffer or one heckuva sophisticated set of spindle-
> motor controllers becomes evident :-)
>
>         The 305 was a stored-program computer, but there were indeed
> quite a few plugboards. One of the most bizarre was the one which
> had hubs which emitted a pulse when a "branch" instruction with a
> particular suffix was executed. That pulse could be routed through
> various tests and then into a portion of the plugboard which
> essentially jammed the program counter.
>
>          _Real_ 305 programmers, please bear with me, or better yet,
> correct me. My knowledge of the 305 comes from reading a manual
> that one of the older guys left around at a buddy's shop. Anyway,
> you would code something like "B*", and a pulse would come whipping
> out of the '*' hole on the plugboard when it executed. You ran wires
> from there through whatever condition you meant to test, such as
> "overflow", "printer on channel 9", "console swith A set", etc,
> in pretty much any combination. Then used a splitter to send the
> resulting pulse into holes for the various values of the various
> digits of the "next pc". Wild. I'd love to get a more accurate
> description, because I'd hate to think _I_ was the only one who
> remembered...
>
>                                         Mike
> | alb...@agames.com, speaking only for myself