Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Re: Ad Hoc SEAint Web Tracking Committee - Subject Coding

[Subject Prev][Subject Next][Thread Prev][Thread Next]
The idea of coding each message for "easy" retrieval assumes that EACH message
contains an otherwise irretrievable gem  (well, I guess you might say that
many of Kate's did but I'm not that sure about the rest of us  :<) ).  From my
observations of the list traffic, many of the messages consist of preliminary
discussions on the way to a consensus.  While they are important in the
context of getting to the final answer, they often don't contribute a long
lasting gem that merits either saving or subsequent retrieval.  Some, in
simple fact, contain information that is not correct.  

Besides, that is what search engines are for.  It would seem far better to
spend time obtaining the "best" possible search engine and organizing the
archives so that search engine can effectively be used (and that doesn't
necessarily require a coding scheme to make it happen.).  

The task you describe seems to require a great amount of volunteer and list
contributor time that might be better spent in solving some of the burning
issues before us (e.g., how do we design a truly ductile steel frame joint,
how do we get a client to pay,  should we really have a dues increase, etc.).

If someone "improperly" codes the message, or it contains information in
multiple areas, categorizization loses its usefulness.  A search engine,
however, can distinguish when multiple topics are discussed and find all
mention of a particular topic.  It can also, using logical search criteria,
limit the of messages one needs to go through.

The archiving of listserver information is much like a paper published in a
journal.  In a paper, whether by a single author or a committee, the thoughts
are hashed over until they are distilled to the salient points.  The incorrect
information, for the most part, is filtered before the information is
published.  Not everything is published or even retained by the author.

IMHO, we should be making the tools available for research efforts, not doing
research or unnessary categorizing that may not need to be done.

My experiences to date with the archives have been pretty good in finding
information  I want.  The search engine could be improved some but it does a
pretty good job (the 4 character limit described in recent posts may need some
work although it has not significantly affected me) as it is.

The beauty of the current list server is that it is simple and useful; it
doesn't require a manual of special codes to use it; and it attracts a lot of
good people from all over the world who are more than willing to share the
benefit of their experience.  I fail to see how "coding" will improve on this.

Bill Cain, SE
Albany, CA