Re: HTTbis spec size, was: Rechartering HTTPbis

On Mon, Feb 6, 2012 at 11:26 AM, Poul-Henning Kamp <phk@phk.freebsd.dk> wrote:
>
> TCP is not "reliable" any more than UDP is, many TCP connections
> fail, both during establishment and later.
>
> What TCP offers is "we'll hide all the retransmissions and pretend
> to be much more reliable than the underlying network" which is often
> convenient.
>

"reliable transport" is a term of art for what you describe above;
I see no reason to redefine commonly understood terms.  At best,
that is going to add time for explanation and at worst, it will add confusion.


>But for many HTTP-purposes, UDP would be good enough:  You send the
>request and either you get a reply or you do not.  That isn't very
>different from a TCP connection breaking.

Personally, I see this as pretty different, but your mileage clearly
varies on that point.

Would you propose different method names for this mode, so that they
were distinguishable
to intermediaries, or would propose that the semantics of the method
be determined by
the transport on which it arrived?

regards,

Ted

Received on Monday, 6 February 2012 20:05:49 UTC