Jonathan Rochkind writes:
 > >> I'm not sure it's a _big_ mess, though, at least for metasearching.
 > >
 > > I wasn't thinking specifically about metasearch, but rather, bad
 > > decisions getting replicated and you end up with an installed
 > > base of bad implementations. The best illustration would be the
 > > huge mess that HTML is.
 >
 > HTML works out pretty well. If our biggest failures were 'failures'
 > like HTML, we'd be doing pretty well.

Got to agree there (even though it undermines the point I was making
before) -- HTML is not a good example of a system that's undermined
its utility by trying too hard to be helpful.

That Clay Shirky observation again: "You cannot simultaneously have
mass adoption and rigor".  It seems pretty clear that it applies to
something like HTML, where you want to have literally millions of
people writing it.  Not so much in implementing search standards,
where the number of implementers is likely in double figures.

 _/|_	 ___________________________________________________________________
/o ) \/  Mike Taylor    <[log in to unmask]>    http://www.miketaylor.org.uk
)_v__/\  "Diagnosing: it is OK." -- wonderful diagnostic from _something_
	 in my AUTOEXEC.BAT