Re: A new suggestion on 100 CONTINUE

>You didn't comment on these suggested changes though and I
>would be interested in your thoughts.

I am not interested in tying HTTP requirements to TCP implementation
details.  I didn't even support introducing all the noise in section 8
(it belongs in an implementation guide, not the protocol spec).
The 100 code was intended to solve an application requirement of not
sending the data before determining whether or not the server requires
encryption or re-encoding of that data.  Since Dave's idea (not yet in
proposal form) would satisfy that requirement, it meets the needs of
the protocol without concern for implementation context.

....Roy

Received on Wednesday, 11 June 1997 14:49:46 UTC