httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From r..@ai.mit.edu (Robert S. Thau)
Subject Re: Stuff that happened at a workshop in Cambridge.
Date Sun, 02 Jun 1996 15:32:14 GMT
  Hmm, I was under the impression that the Catalog servers were using the 
  Harvest model and technology... 

They are using the Harvest model, and *portions* of the Harvest
technology (in particular, SOIF) --- however, they are *not* using
the Harvest gatherer/broker protocol.  Instead, they're piggybacking
the requests over normal HTTP, in a fashion which, by design, could
be implemented on J. Random Server by installing a CGI script (or so,
at least, was the sales pitch which one of the ex-Harvest guys
delivered).

> 5) Yet *another* thing that may appear over the medium term is a
>    common full-text inverted index format.

  I though that's what SOIF was... ah well.  

SOIF is a common format for attribute/value associations and other 
metadata (author, title, keywords, and so forth).  However, to the
best of my knowledge, it is not really suitable for exchanging 
full-text indexes of entire sites.

(FWIW, the Stanford effort that I mentioned in the original note is
considering using SOIF as the exchange format for queries and responses,
in the same role for which straight Z39.50 uses BER --- the difference
being that SOIF can be parsed in a couple of pages of C code, and less
than that in any really decent language).

rst

Mime
View raw message