Re: WGLC draft-ietf-webdav-redirectref-protocol-12

This is a very simple subject area, but there are some
choices to be made, and without a standard to define a common way,
we'll get folks just rolling their own with incompatible semantic
details (as is being done now).

So unless there are some actual technical problems with the REDIRECT
draft, I believe the current REDIRECT draft should go to "Proposed",
to address the interoperability in this area that already has arisen
due to the lack of a standard for how to author redirect references.

In this case, since there aren't any subtle technical issues/obstacles
to be addressed, and we just need a common convention, I think it is
appropriate for the draft standard to drive the implementations, rather
than the other way round.  In particular, the issue of authoring redirect
references isn't critical enough for folks to modify their current 
mechanisms to make them interoperable, until there is a draft standard
that defines how to do so.

Cheers,
Geoff


Julian wrote on 06/14/2005 04:38:11 AM:

> 
> Lisa Dusseault wrote:
> > 
> > 
> > The main flaw in the REDIRECT proposal is that there is not enough in 
> > the way of plans to implement it.  Without a set of independent 
> > implementors around to review it, I fear it's too complex or has 
missed 
> > key interoperability issues.
> 
> I do agree that there seems to be more interest in BIND than REDIRECT. 
> On the other hand, I disagree that because of this it can't start at 
> "Proposed" (as Jim W. pointed out some time ago in a similar dicussion 
> for BIND).
> 
> If the working group feels that it's too early to go to "Proposed", I 
> think "Experimental" would make sense. It would preserve the work that 
> has been done; and if at a later point of time more implementations 
> appear, it can be rev'd up.
> 
> > To be clear, I do understand that the Web needs and uses redirects, 
and 
> > I see that administrators do create them and that browsers follow 
> > redirect status codes.  I'm arguing that there isn't a clear need for 
> > interoperable authoring of redirect resources, or if there is, it's 
not 
> > met by this specification.  Implementors might tell us, for example, 
> > that they don't need the ability to modify a redirect (why not just 
> > recreate) or that they'd prefer something which could handle 
redirecting 
> > URLs via pattern matching to another set of calculated URLs.
> 
> The ability to modify was added based on a Last-Call comment in 2000, 
> and the WG decided to accept it. See 
> <http://greenbytes.de/tech/webdav/draft-ietf-webdav-redirectref-
> protocol-issues.html#lc-58-update>.
> 
> > There seems to be one organization that implements REDIRECT authoring 
in 
> > a potentially-interoperable way -- I believe that's Julian's 
> > organization. I think that's great, and even better that they're 
> 
> It's SAP (first time I've heard it called "Julian's org", but that 
> sounds nice :-). As a matter of fact, the Xythos client implements some 
> aspects of the protocol as well (discovery but not authorability).
> 
>  > ...
> 
> Best regards, Julian
>