29 May
2005
29 May
'05
5:02 p.m.
On Sat, 28 May 2005, Ulf Andersson wrote:
The earliest reference I have found to a working drum
memory is the
April 1949 version of the University of Manchester (UK) Mark I. This
is said on page 114 in the little book "Early British Computers" by
Simon Lavington. Anyway, that is 56 years ago.
The problem with histories is that they oversimplify. In this
instance it's not too bad; RAMAC is a disk drive by modern
definitions, so there's a reasonably bright line between it and
similar things.
But things like Booth's paper memory defies modern categorization;
though rotating, magnetic, disk-like, it was a main-memory for a
non-stored-program calculator. Since in 1947 there were no
computers-as-we-know-them-today -- meaning an electronic machine
with an "operating system" and a a data abstraction known as a
"file" -- this disk/not-a-disk doesn't fit into "timelines" and
Famous Person/Famous Product histories, so they get dropped.
This problem exists in other areas of history and historiography,
but in most areas of human culture -- clothing, food, language,
housing, warfare, artwork, etc -- massive paradigm shifts don't
occur every 15 minutes.