RE: drop QSD?

At 02:59 PM 4/16/00 -0700, Bernard Chester wrote:
>Jim:
>
>As we discovered in DMA, without this you cannot implement a decent
>interface independent of prior knowledge of the service.  Maybe you want to
>address the complexity rather than dumping it?

Indeed, this was the reason I put it in, and, as is doubtless obvious to
you or anyone familiar with DMA, the design owes a great deal to it.

Yet so far as I know, neither of the two implementations of DASL implements
it, and if it is in fact true that nobody WILL implement it, we don't do
any good by having it in the spec.  it will just get removed for us later on.

I don't actually think the QSD spec is all that complex.  But I also have
been led to see that my threshold for complexity is different from that of
the mainstream.  And the IETF rules are that implementations matter, not
spec-writer's opinions of elegance.  So if no one implements it, it has to
go, even if its dead simple.

Received on Sunday, 16 April 2000 19:12:37 UTC