Well. It may be work hours but since I have all of my desktop machines and
the farm thrashing away and computing as hard as they can I find myself
with a little bit of free time to write up a post.
Have you noticed that no matter how much faster the computer gets, or
how much hardware you buy or use, you’re still stuck waiting for the computer?
As it is, these "piteously slow" dual gig chips are whimpering
with the work I am throwing at them. You think we’d be able to do more than this.
Or maybe things aren’t that simple.
Way back in the very early days of computing. And we’re talking about the days when
Howdy Doody and Mickey were on the TV, computers ran slow too.
One of the first successful large scale computers was made by the Unisys corporation,
called the UNIVAC. Only 46 UNIVACs were ever produced. Still, they changed the world.
Among the many firsts the UNIVAC had were things like the first multimedia
presentation (a UNIVAC played "Anchors Aweigh" at it’s own dedication for the
Navy’s Applied Mathematics Lab in 1953) and the first computer game (NIM, created
in 1953 that no human has ever beat the UNIVAC at) as well as many others.
Probably the most famous UNIVAC moment was when CBS television used one to predict
the 1952 election. Not believing the computers prediction based upon a voting sample,
CBS didn’t air the results for nearly an hour. When they did, they were forced to admit
that the computer had been very much correct. (Ike won, by the way) It’s a famous moment
but I’m not really talking about arguing the integrity of computer processed data, more
about the processing time.
Receiving less note was a task performed in 1956. The Franklin Life Insurance company
donated use of it’s UNIVAC to John W. Ellison. John, a rector in the Episcopal church of
Massachusetts, had been tasked with assembling the concordance for the Revised Standard
Version of the Bible. (A "Concordance" is an alphabetical index of all the words
in a text, showing every contextual occurrence of a word).
A staff of fifteen persons entered the entire text of the bible via punch cards and
UNITYPERs over a period of 6 months. Two complete entries were created so that they could
be compared to locate typographical errors from entry. This produced 10 reels of output.
Then the processing occurred. The main keyword extraction took 250 hours of processing.
The final alphabetical sort of the data took another 50 hours, producing the output set
of 26 reels.
So 300 hours of run time on a computer the size of a garage. So what you may think? Why
is that so special.
The previous concordance was assembled in 1894 and had took 30 years to perform by hand.
Maybe the computers aren’t so slow anymore. Maybe we’re doing a lot more with them. Think
about this. When you throw away that musical greeting card you throw away more electronic computer
than existed on the face of the earth in 1945.
Would you look at that. My processing is done. Back to work.