Re: [Bug 188] PROPFIND include-dead-props

Lisa Dusseault wrote:
> 
> I finally paged in enough context to be able to consider this set of 
> changes.
> In particular, the justification, from bug text:
> <http://ietf.cse.ucsc.edu:8080/bugzilla/show_bug.cgi?id=188>
> 
> "The difference here is that with the originally proposed extension, the 
> client
> can tell whether the server evaluated the extension, while with the 
> definition
> in RCF2518bis, it can't (it has no way knowing whether the server just 
> ignored
> the extension, or happens to have no dead properties)."
> 
> That's not quite true, the server MUST support dead-props if it advertises
> support for 'bis'.   Besides, a smaller fix could be for the server to 
> return an
> extra element saying "no-dead-props".

But if a server implements "bis", it MUST also support lots of other 
unrelated features. This is a question of granularity, and optimally, we 
won't need "bis" at all because all the things we add can be discovered 
individually (such as support for DAV:lockroot, for example).

> That said, has anybody at all implemented this?  If not, then I guess 
> the change is
> harmless even if the justification isn't quite true.

We have implemented it but found we're not using it in our own client 
for the reasons given above. So the new revised proposal isn't based on 
theory, it's based on running code. This is the approach we find is 
working, and we've been using it since 2002 when we initially proposed it.

Best regards, Julian

Received on Saturday, 3 December 2005 23:17:35 UTC