On Tue, Apr 06, 2004 at 05:27:41PM -0500, Jason E. Sibre wrote:
> I still prefer something a bit more explicit than the tuplized 'mapping'.

+1 with bells on.

> No Magic, and all...  Anyway, it's not likely to hurt me as long as
> the old behavior continues to work, though I do wonder slightly about

I somehow managed to lose a lengthy reply that addressed this, but
perhaps it's better this way.  I think that if the function of
_q_exports is to be changed, it should be a big step; then a desire for
compatiblity seems to push towards the introduction of, say, _q_map, a
dictionary that explicitly maps from URL components to... well, that
might be a separate discussion, as might the question of whether
mapping is attempted only once or if it can chain.

But I think this is much saner than piling magic special cases on
_q_exports.  Should be easy to add _q_map (if it exists) prior to
trying the old _q_exports to allow for an easy transition, then a quiet
obsolescence (assuming there's agreement that the _q_map equivalent of
_q_exports is simple enough to allow that).

Oh, and then there won't need to be any more discussions about the
overhead of looking up URL components even when the lists get long. :-/

--
Education makes people easy to lead, but difficult to drive;
easy to govern, but impossible to enslave.  -- Henry Peter Brougham