> Date: Mon, 15 Sep 2003 10:29:49 -0400
> From: Ray Denenberg <[log in to unmask]>
>
> I like the proposal [...]

Excellent!

> [...] but I would like to strongly suggest:
> 1. Continue to call index sets "index sets", and call their prefixes
> "index-set prefixes".
> 2. Call the new prefixes, for relations and relation modifiers, "
> relation prefixes", and "modifier prefixes".

Er.  In part (1) of this, did you mean that we should call them "index
prefixes"?  It seems odd to say "index-_set_ prefixes" and not
"relation-set prefixes" or "modifier-set" prefixes.

I think the point here (and the reason we felt the name needed to
change) was that these babies are no longer just sets of index-names,
but sets of ... well, all sorts of context.

> I'm suggesting this so that we completely disabuse ourself of the
> illusion that we're recreating the notion of attribute vectors.

Eh?  Why would that be an illusion?

> An index prefix, relation prefix, and a modifier prefix could all be
> the same string.

Precisely -- they'd all be the same string because they're all from
the set, even though they're three different kinds of thing.  Which is
why, again, the name "index set" no longer seems to serve.

> (Aside: what you're calling "modifier", cql currently calls
> "qualifier".  Which do we want to call these?)

This is partly my fault, I agree: when the rest of the world was
calling indexes "indexes", I was still calling them "qualifiers".  I
still like my name better (goes without saying :-) but I've given up
since consistency is more important than correctness.  So now we're
all calling indexes "indexes", but I have polluted the term
"qualifier" by using it in this way.  (See the tutorial, which still
talks in these terms).  So I think this is an unsafe term, and we
should not use "qualifier" at all in CQL discussion, instead sticking
with "modifier" -- which has in fact been used pretty much universally
except in the grammar on the MA site.

So I'll do a deal with you :-)  Change the grammar to talk about
modifiers, and I'll fix the tutorial to talk about indexes.

> (Minor editorial suggestion: could you make first and second
> rejected solution 1.1 and 1.2 instead of 2 and 3?)

Done.

 _/|_    _______________________________________________________________
/o ) \/  Mike Taylor  <[log in to unmask]>  http://www.miketaylor.org.uk
)_v__/\  "Out of all of the people that ever knew, there should have
         been one I could believe in -- and it should have been you"
         -- The Waterboys, "It Should Have Been You"

--
Listen to my wife's new CD of kids' music, _Child's Play_, at
        http://www.pipedreaming.org.uk/childsplay/