Re: translation for techno-peasants

[Kevin Cabral <kcabral@xxxxxxxxxxxxxxxxxxxxxx>]

On Tue, 23 Jul 1996, Jon Beasley-Murray wrote:

> Kevin recently posted this to the marxism l*st, and I thank him heartily
> for it, to enable people like me to figure out what's going on.
>
> I do have a question, however: I suspect that there are a variety of ways
> of constructing an archive search engine.  I personally am truly glad
> that we are considering installing one.  Indeed, I personally think that
> any search engine is better than none.  However, even looking at the main
> internet web search engines I see that some are much more handy than
> others... most often, some seem to get tiny details which are of little
> use while others miss the big obvious stuff.  I hope that we can do a bit
> of reviewing the options, perhaps--if we decide to go ahead with this--to
> work with Kevin on how it will all work.

        Sure, please give suggestions. Basically the search engine I have
thought of will be able to accept search criteria both in a simple boolean
language (using terms like and, or, if, and then in defining a search) or
accept queries by Author name, Date (or Month, Year), Subject (what
appears in the header) or a combination of those search criteria. If
enough volunteers came through we could index each message with a set of
keywords that would be consulted first, some engines on the Internet such
as Magellan do this. They read a message (or in their case a page) and
assign descriptive words to it. For example a post could be designated as
Lenin, Althusser, and Freud if it gave signifigant information on all of
them; or Louis' series on Nicaragua could be Nicaragua, land reform, and
peasantry.

        Physically the search engine would search through the body of each
message (or a header, or special area inserted to designate special
"descriptor terms") for the text. This can be done fairly quickly,
probably in less than a minute using Jefferson (what I'm calling the
computer at University of Virginia that houses the archive) remotely; or
even faster by executing the program, and storing the data files on your
own computer. Either way it won't be as fast as a hardware and memory
mammoth like AltaVista (altavista.digital.com) but it should be
acceptable.

Kevin
Cols, Oh





     --- from list marxism@xxxxxxxxxxxxxxxxxxxxxxxxxx ---