This type of storage was originally suggested by the Eckert/Mauchly
team at the University of Pennsylvania, who built the Eniac and used
delay lines themselves in their commercial follow-on, the Univac. One
problem with it: when cracks developed in the quartz crystals, it
wouldn't affect their response to vibration. A crack could thus
propagate in the crystal without affecting the function of the memory
at all until it reached the point of physical failure, at which point
you'd have a shattered crystal and a tube full of mercury on the
floor. On the whole, I'd rather have SDRAM SIMMs.
> Ivars Peterson's Mathland, "Computing with the EDSAC"
> > Wilkes recalls in his memoirs, "By June 1949, people had begun to
> > realize that it was not so easy to get a program right as had at one
> > time appeared. I well remember when this realization first came on
> > me with full force. The EDSAC was on the top floor of the building and
> > the tape-punching and editing equipment one floor below.... I was
> > trying to get working my first nontrivial program.... It was on one of
> > my journeys between the EDSAC room and the punching equipment that ...
> > the realization came over me with full force that a good part of the
> > remainder of my life was going to be spent in finding errors in my own
> > programs."
The Cambridge lab's manual for users of the EDSAC, by Wilkes,
Wheeler, and Gill, was published in the United States by Addison,
Wesley, and is one of the first books ever published on computer
programming in the modern sense. (Possibly the first, depending on
whether you count von Neumann's widely circulated papers, or the ETA
book on the construction of computers generally). This book was
reprinted a few years ago as part of the Babbage Institute's reprint
series; the original is quite rare, but it probably isn't hard to find