W3C home > Mailing lists > Public > xml-dist-app@w3.org > February 2002

Re: Rules for intermediaries handling of PIs, whitespace, etc.

From: Mark Nottingham <mnot@mnot.net>
Date: Fri, 1 Feb 2002 06:51:12 -0800
To: Joseph Hui <jhui@digisle.net>
Cc: Don Mullen <donmullen@tibco.com>, xml-dist-app@w3.org
Message-ID: <20020201065111.C1168@mnot.net>


I'd even settle for a canonical form of section 5...

Cheers,



On Thu, Jan 31, 2002 at 12:22:26PM -0800, Joseph Hui wrote:
> Equally important to defining the rules and yet much more challenging
> is defining a viable mechanism (or at least the requirements for such
> mechanism) to enforce the rules.  I for one would like to see W3C
> accomplish what IETF couldn't in dealing with intermediaries in ways
> that can *safely* allow for intermediary processing -- which inevitably
> results in altering data byte wise -- without setting off yet another
> end-to-end religious war.  
> 
> On defining rules, perhaps we should strive to achieve an ultimate
> goal: a standards rule set to render any given SOAP document in a
> canonical form.  E.g. parserA and parserB may exercise different
> features on a SOAP doc, but they MUST produce the same output if
> the rule set kicks in.  
> 
> Regards,
> 
> Joe Hui
> Digital Island, a Cable & Wireless company
> ===================================================
> 
> 
> > -----Original Message-----
> > From: Don Mullen [mailto:donmullen@tibco.com]
> > Sent: Thursday, January 31, 2002 6:47 AM
> > To: xml-dist-app@w3.org
> > Subject: Rules for intermediaries handling of PIs, whitespace, etc.
> > 
> > 
> > 
> > During the conference call yesterday, I took an action to 
> > start a discussion
> > of the SOAP 1.2 rules that should be defined for describing what
> > intermediaries are allowed to do with PIs, lexical space, 
> > encodings, etc.
> > 
> > Some items to consider during discussion:
> > o white space (including line terminators)
> > o processing instructions
> > o comments
> > o encoding
> > o character reference preservation (or introduction?)
> > o <foo/> vs. <foo></foo>
> > o order of attributes
> > o other items considered outside the Infoset? (see Appendix D 
> > of Infoset
> > spec)
> > 
> > Options include:
> > 1) Allow making lexical changes anywhere
> > 2) Disallow making lexical changes anywhere
> > 3) Disallow lexical changes within a block, allow elsewhere
> > 
> > Don Mullen
> > TIBCO Extensibility
> > 
> > 
> 

-- 
Mark Nottingham
http://www.mnot.net/
 
Received on Friday, 1 February 2002 11:38:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:06 GMT