Re: Bugzilla issue 10, was: Plan to resolve issues in 2518bis

Lisa,

Based on my reading of the book "Effective XML"
(http://www.cafeconleche.org/books/effectivexml/)

Item 10: White space matters

  WebDAV servers should preserve the white spaces
  when the attribute "xml:space" is set to "preserve".
  Else, if "xml:space" is not set or set to "default"
  "the application may treat white spaces in the
  element in whatever fashion is customary for that
  application".

Item 21: Rely on namespace URIs, not prefixes

  A WebDAV server should be allowed to change any
  namespace prefix.  "It's the URI that counts,
  not the prefix".

BTW, I highly recommend reading "Effective XML" to
anyone that works with XML.

Cheers,
Bernard

Lisa Dusseault wrote:
> I guess what's giving me so much cognitive dissonance here is that 
> prefixes now are not preserved but whitespace is. That seems 
> inconsistent to me -- if some XML rewriting is OK but other XML isn't, 
> what's the difference.
> 
> Lisa
> 
> On Oct 13, 2005, at 12:36 PM, Julian Reschke wrote:
> 
>     Lisa Dusseault wrote:
> 
>         On Oct 13, 2005, at 12:07 PM, Julian Reschke wrote:
> 
>             Actually what I'm asking for is that we don't change the
>             text unless there clearly is a consensus to do so. The
>             current spec says whitespace is significant, and as far as I
>             can tell, nobody has asked for a change of that.
> 
>         Fair enough, although now I'm thinking we should be specific
>         about XML values, and about the beginning/end of the value as
>         well as the middle. I was thinking that the existing text in
>         2518 applied only to text property values.
> 
> 
>      From <http://www.webdav.org/wg/rfcdev/issues.htm>:
> 
>     -- 
>     107
> 
> 
>     IS_XMLSPACE_SIGNIFICANT
> 
> 
>     Edit
> 
> 
>     InBis
> 
> 
>     Should the xml:space attribute be respected.╴ 2518bis on 6/1/02 says
>     it should not. There is some debate on this.
> 
> 
>     Re-raised:
>     http://lists.w3.org/Archives/Public/w3c-dist-auth/2002AprJun/0137.html
> 
> 
>     The conclusion of the May/June 2002 discussion was that white space
>     is significant:
>     http://lists.w3.org/Archives/Public/w3c-dist-auth/2002AprJun/0152.html
>     -- 
> 
>     So the old issues list says this was discussed, that there was
>     consensus, and that rfc2518bis has been changed accordingly.
> 
>     If you want to re-open the issue, please do so (but in a different
>     thread). If you do, please make sure to clarify what was wrong the
>     resolution we reached back then.
> 
>         I'm still very curious to hear what implementations do/assume;
>         that's good input to see if the consensus is consistent with the
>         spec.
> 
> 
>     Do you have any data about servers that get that wrong? That would
>     be interesting indeed.
> 
>     Best regards, Julian
> 
>

Received on Monday, 17 October 2005 21:20:45 UTC