> Date: Mon, 17 Jun 2002 11:50:27 -0400
> From: Ray Denenberg <[log in to unmask]>
> > > We explored all of this, in great depth, at our first meeting a
> > > year ago. There was strong sentiment not to do this as it would
> > > be a futile effort to define negotiation that wouldn't work in
> > > practice anyway.
> > Why not?
> This is my recollection of the thinking when we discussed it a year
> ago: servers in general aren't going to be flexible enough to alter
> their behavior based on the user request, nor even to reject the
> request in advance.
I don't understand why we should compromise our protocol for the
benefit of such broken servers.
> In other words, optimally, if you had such a flag, you'd want the
> server to make the result set persistent if you asked it to; but
> lacking that, at minimum, you'd want the sever to tell you if it
> cannot honor the request. The thinking was that as a practical
> matter servers can't even do the minumum, so a flag would not only
> be useless, it would be misleading.
Then these misbegotten piles of bits are simply not worthy to be
called SRW servers, surely? We are _defining_ the protocol here.
Let's not give up on getting it right before we've even started.
/o ) \/ Mike Taylor <[log in to unmask]> www.miketaylor.org.uk
)_v__/\ "But I remember ... Three lions on the shirts; Jules Rimet
still gleaming" -- Skinner, Baddiel & Broudie, "Football's
Ah, how appropriate.