Re: Client-side searching proposal

Rob Hartill ([email protected])
Thu, 26 Jan 1995 18:12:45 +0100



John Granks wrote..
>
> But an interesting thing occurred to me about the nature of WWW while
> thinking about the relative merits of client side versus server side
> implementations for this kind of functionality. If it is done on the
> server (as it is in WN) then for that server it just works -- with every
> client. On the other hand to do it in the client requires the daunting
> political task of arriving at a consensual standard and getting it
> implemented in all the browsers.

The original proposal, which has been refined through suggestions here,
doesn't require implementation on all browsers. Browsers which support
it will be more attractive than those which don't. Browsers choosing
not to implemement it will continue to work, just not as effectively
as those which do implement it. The "daunting political task" seems
to be happening right now, and appears to be running smoothly.

To recap the current state of the suggestions as I understand them is..
Opting initially to get just the basics of the idea implemented, we'll
use the anchor character "#" (for backwards compatibility), followed by a
special character or word, "!" or "find" (or whatever), followed by an
"=", then a keyword or comma separated list of keywords,

e.g.

http://x.y.z/abc#find=dog
or
http://x.y.z/abc#find=dog,cat,pig

Browsers can choose to implement the search in whatever way they like,
the simplest could be to look for "dog" - ignoring "cat" and "pig".
In the best implementation, the browser would find all occurrences of all
the words and allow the user to move freely from one to another.

The use of "find=" will allow other types of client side (or server side,
if appropriate) features to be added as and when they are proposed and
agreed upon. e.g. if someone wants regular expressions then "reg=" might
be used.

rob.

--
Robert Hartill                           http://nqcd.lanl.gov/~hartill/
Los Alamos National Laboratory                    Phone: (505) 665 2280
Theoretical Division, T-8, MS B285                  Fax: (505) 667 5585
P.O. Box 1663
Los Alamos, NM 87545, U.S.A.