Re: Idempotence

In general folks, (Maurizio in particular in this case), if
there is wording in the specification that you believe can be clarified
to prevent mis-understanding of the HTTP specifications
as people implement it, now is the time to complain about it, 
so we can fix it for draft standard, and prevent future problems.  

If another sentence, a rewrite of existing text, or better 
cross-referenceing would help, I certainly want to know about it.
The primary authors of the document have read the document until 
we no longer actually see problems; it is only
by questions and problems being taken all the way to suggested
improvements and clarifications of the document that further improvements
can be made.  All of us who wrote it are burned out on the words.

So, for example, if a definition of idempotence in the glossary
might have helped in understanding this issue, we can certainly
add such things.  If there are redundant words in the document confusing
something, we can delete the redundancy.

Editorial improvement of the specification is certainly allowed
(and encouraged) in the period between proposed and draft standard;
if we need to fix wording (or of course find the protocol itself
is fundamentally flawed in some way), we can and should fix the
problem.  Some IETF documents get entirely rewritten in this period
(though I am NOT volunteering to do a massive rewrite in this case;
I will guide my choices by the doctor's mantra "first, do no harm".)

Your editor,
				- Jim Gettys

Received on Thursday, 26 September 1996 07:35:48 UTC