Re: twinql Retrospective motivates DESCRIBE refinement?

On Tue, 2005-09-27 at 09:57 -0400, Kendall Clark wrote:
> On Tue, Sep 27, 2005 at 08:33:25AM -0500, Dan Connolly wrote:
> > 
> > Did we consider this design already?
> > 
> > 
> > >    As I write, it occurs to me that some way to say which method is  
> > > being (server -> client), or should be (client -> server), used for  
> > > DESCRIBE would be desirable -- I'd like my clients to know that  
> > > they're getting CBDs, or the clients might wish to ask for a certain  
> > > kind of description.
> >
> > 
> > 
> > I'm considering re-opening issues#DESCRIBE. Advice is welcome.
> Is this orthogonal or related to SADDLE? I would have thought such
> description (the server does *foo* when the client says DESCRIBE) would have
> been a SADDLE issue, though I guess the flip side is that the client says to
> the server "I want you to do *bar* when I say DESCRIBE".

Yes, issues#serviceDescription is also possibly relevant.

My question is the same in either case: did we consider this already?
I think I did; it's part of the reason I object to DESCRIBE.
But I'm not sure I can confirm from records that the whole WG
considered such designs.

Does anybody think that it's new information that they would
like to use to change or reconsider the WG's decision?

> Cheers, 
> Kendall
> --
> Sad songs and waltzes aren't selling this year... --Cake
Dan Connolly, W3C
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Tuesday, 27 September 2005 14:04:22 UTC