Earlier microfilm retrieval systems (Was: What is this?

[Fred Cisin]

On Fri, 10 May 2019, Chuck Guzis via cctalk wrote:
> Reading about WALNUT, it was more than a little unusual for its time.
> The idea was the setup stored (photographically) almost a million images
> using a non-silver process.  The images were indexed digitally and the
> index was searchable.   The output appears to be a standard  aperture
> card.  Although both of the references that I found mention
> Kalfax/Kalvar media, WikiP says that the systems delivered to the CIA
> used a different diazo process that was apparently more stable than the
> Kalvar process.

There have been quite a few systems for computerized retrieval of 
photographic images.
Besides "aperture cards" there were a number of earlier systems that 
encoded an ID on the microfilm.

Lest we be overly concerned with that using too much of the microfilm 
physical space, keep in mind that soundtrack was often included on 
16mm and 35mm movies.  Maurer, and others?, had up to EIGHT parallel 
analog audio tracks in the margin of movie film.  Eight bit parallel!?!

90 years ago, Emmanuel Goldberg built a system:
http://people.ischool.berkeley.edu/~buckland/statistical.html
http://people.ischool.berkeley.edu/~buckland/goldberg.html
http://people.ischool.berkeley.edu/~buckland/goldbush.html
One of several reasons why Goldberg is largely unknown is that he was 
working in Germany (Zeiss) when the Nazis came into power, and Zeiss 
had to systematically wipe the records of existence of Jewish engineers.
His optical reading of the metadata was done by "complement" or 
"extinction".


Vannevar Bush's speculative article "As We may Think" (Atlantic Monthly 
1945, although first written in 1939) suggested possibilities for the 
"Memex".
It used "coindcident" templates, instead of "extinction" for the matching. 
It also talked about the microfilm remaining in motion, with a Xenon 
flash tube to "stop" the motion.  The level of magnification and the 
speed that he talked about moving the film were not consistent with the 
speed of the flash tubes that were actually AVAILABLE at the time, nor in 
the 1950s.

Vannevar Bush did not reference the work of predecessors, such as Goldberg 
and a few others, although there is some evidence that he had at least 
heard about Goldberg's work.  Atlantic Montlhly was popular press, not a 
"peer-reviewed" academic journal, so policies of citations were lax.

Bush seems to have not been a believer in hierarchical information 
storage, and used examples of flipping from documents to other documents 
in "trails".
Ted Nelson credits that with inspiration for "hypertext".
Berners-Lee credits Ted Nelson's hypertext with inspiration for WWW.

"As We May Think" has caused Bush to often be considered the "father of 
Information Science", which frankly, I don't feel is much more accurate 
than some of the modern crediting of Steve Jobs and Bill Gates with 
"inventing" the computer.