Question (complaint?) on CLOS Spec

[Mike Thome <mthome@BBN.COM>]

(I'm referring to 88-002R of the CLOS spec, so if these (bugs?) have been
fixed since this version of the spec, please ignore...)

In the discussion on define-method-combination, I've got a few questions
and complaints on the spec:
	1. What was the reasoning behind having short-form-defined method
combination styles have thier primary methods defined as those with
a qualifier the same as the name of the combination style?  Clarity?
	2. The examples given in this section make no use of defgeneric -
which I suspect must be required, since I see nothing in the spec that
prohibits the name of a short-form-defined method combination from being
the same as one of the optional qualifiers for a long-form-defined
combination style.
	3. Generally, it seems inconsistant that standard method
combination has null method qualified primary methods, while all the rest
of the examples (as well as any short-form-defined combinations) have
non-null primaries.

curiously,
	-mike thome (mthome@bbn.com)