Fwd: Re: formatting and addresses - Hodges
Bill Wood
pmmail@rpglink.com
Tue, 06 Jul 1999 09:32:49 -0700 (PDT)
On Mon, 05 Jul 1999 08:39:38 -0400 (EDT), Darin McBride
wrote:
>On Mon, 05 Jul 1999 10:49:05 +0100, Paul Hodges wrote:
>
>>The standard explicitly says that implementations must be able to
>>receive this length but "must not send" longer, but the next sentence
>>(in capitals, and repeated lower down) urges that implementation
>>techniques should be used that do not inherently force this
>>limitation. This qualification implies a desire to remove the
>>(strictly mandatory) limitation of the previous sentence, but nothing
>>has apparently been done about it.
>
>The qualification states to me the normal internet standard: be
>conservative in what you send, and liberal in what you receive. Normal
>stability requirements.
>
Well, yes and no. While it is desireable to minimize
traffic on any network to what is needed, the bandwidth
on the Net is so gigantic relative to 1000 byte segment
that it is truly inconsequential, and, I'm rather sure,
traffic minimization was not the purpose of the
standard.