W3C home > Mailing lists > Public > www-zig@w3.org > August 2003

Attribute Architecture -- new type?

From: Ray Denenberg, Library of Congress <rden@loc.gov>
Date: Fri, 8 Aug 2003 14:08:43 -0400
Message-ID: <024101c35dd8$1abed920$849c938c@lib.loc.gov>
To: "ZIG" <www-zig@w3.org>

Back to the attribute architecture problem (after a few weeks away) --

 I  propose we define a new "type", along the lines that Alan was originally
suggesting.

Let's start with two points we all agree upon:
1. allWords, anyWord, adjacentWords need to be changed from structure to
comparison attributes.
2.  We need to distinguish between word and string indexes.

And of course it's the second of these that's interesting,  exemplified by a
search on title --  are you looking for the supplied words or the entire
string within the title? Or, as Ralph once said  (in a post to SRW): "Word
indexes are indexes that support implicit proximity between a  provided list
of words.  String indexes are indexes that require exact matches on the
entire content of the supplied search term."
 I'm uncomfortable using format/structure for this distinction. As Alan
suggested, look at the the bib-2 format/structure attributes: each is
defined for a specific access point type. For example, "iso 8601" is a
format/structure value, defined for use when "duration" is the access point.
(Wouldn't make much sense if the access point were "Map Scale".)

Now let me digress for a moment. I think the discussion so far has been
encumbered by the presumption that any solution must be consistent with the
attribute architecture as currently defined. I'm challenging that
assumption. If someone comes forward and objects claiming an investment in
the attribute architecture then perhaps I'll back off of this challenge. If
nobody makes that claim, I don't see why we can't make some fundamental
change, if we think it's crucial. I don't want to open it up for major
revision --  many people think  that the architecture is a good piece of
work -- but if we can easily improve it, let's do it.

Anyway, it seems to me that the "indexing method" is orthogonal to the
format/structure, at least as format/structure is envisioned by the bib-2
usage (another good piece of work, in my view, and the only effort to date
that's based on the arhitecture).

I suggest that we add a new type. If you go back to the beginning of this
thread, Alan hinted that  he really wanted to suggest a new attribute type,
along the lines of indexing-  or term-extraction method, but that he felt
such a suggestion would be rejected as too radical.  I think it's a good
idea.

--Ray
Received on Friday, 8 August 2003 14:08:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 29 October 2009 06:12:23 GMT