> Date: Fri, 7 Jan 2005 17:25:22 +0100
> From: Theo van Veen <[log in to unmask]>
>
> This is a typical example of an index that is choosen by the user
> when the user doesn't which index to choose or doesn't care about
> the actual index. We cannot expect that a user is able to decide
> which one to choose from serverChoice, anywhere, anyknown,
> anywhatsoever given the discussion on this item.  So I want to
> propose that if *one* of those indexes is available that there
> should be *at least one* index name that a client can count on
> without having to search in explain for subtle differences in
> serverChoice, any, anywhere etc.

(I know you won't believe this, Theo, but ...)

I agree with Theo!

:-)

I do think the distinctions between these various special indexes are
potentially important and well worth the effort of getting right; but
I also think we need to make sure that people who are throwing
together quick clients don't have to think about these subtleties
unless they want to.  It seems obvous to me that the hidden subtext of
serverChoice is "Just Do The Right Thing, Dammit!", so whatever other
indexes we end up adding to the "cql" set, I suggest that we have a
big, bold banner saying "If you don't know or care which of these to
use, just use serverChoice".  (Which of course is equivalent to not
specifying an index at all.)

 _/|_    _______________________________________________________________
/o ) \/  Mike Taylor  <[log in to unmask]>  http://www.miketaylor.org.uk
)_v__/\  "Things like xmlrpc, soap and the stuff on top of them are
         designed to 'interwork through firewalls'.  A better phrase
         would be 'go through the firewall like a knife through butter
         in a way that prevents the companies involved monitoring
         the activity'" -- Alan Cox.

--
Listen to free demos of soundtrack music for film, TV and radio
        http://www.pipedreaming.org.uk/soundtrack/