Re: Who can express URL syntax with BNF

Tony Sanders <sanders@BSDI.COM>
Errors-To: listmaster@www0.cern.ch
Date: Sat, 30 Apr 1994 10:59:39 +0200
Errors-To: listmaster@www0.cern.ch
Message-id: <199404300856.DAA00885@austin.BSDI.COM>
Errors-To: listmaster@www0.cern.ch
Reply-To: sanders@BSDI.COM
Originator: www-talk@info.cern.ch
Sender: www-talk@www0.cern.ch
Precedence: bulk
From: Tony Sanders <sanders@BSDI.COM>
To: Multiple recipients of list <www-talk@www0.cern.ch>
Subject: Re: Who can express URL syntax with BNF 
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Organization: Berkeley Software Design, Inc.
Organization: Berkeley Software Design, Inc.
Rick Troth writes:
> > The server should use the URL to disambiguate 
> > whatever it needs to disambiguate: 
> 	I disagree. 
> > 	GET /place-order?...	form
> 	It's up to the object.   Let the server(s) disambiguate 
> based on the attributes of the object being GETted. 

Then you don't disagree.

I didn't say *how* you decide that "/place-order" is a form, I just said
that you do.  I don't care if you hardcode it, look up some attribute,
call the GET method of the object, or ring up your mom and ask.

All of those are viable implementations (although the latter has some
nasty delay proporties).  Most (probably all) of the servers (certainly
CERN, NCSA, and Plexus) currently use some form of #2 (like you suggest)
but #3 is actually the superior implementation (because it can easily
emulate 1, 2 and 4).  I am working on converting Plexus to #3-style in my
copious free time.

--sanders