On Tue, 16 Jun 1998 22:36:30 -0400 Rich Kulawiec <[EMAIL PROTECTED]> said:
>Hmmm. Given that you make software for the Internet, don't you have
>someone read ALL the draft RFC's? If not, why not?
For the same reason that we don't have someone reading ALL the Microsoft
technical articles about Windows NT, ALL the Sun technical articles about
Solaris, and so on. Which incidentally are a LOT more relevant on the
average than draft RFCs. Mailing lists are not standardized by the IETF
and will probably never be, and as for the MIME and SMTP RFCs, we follow
them of course, just as we follow selected NT issues and are a field test
site for many operating systems, which doesn't mean we read articles
about COBOL compilers or SNA.
>The -request convention has been around since the Arpanet. I remember
>seeing it back in 1979.
Yes, and it pointed to a human person. LISTSERV supports the -request
address as (by default) an auto-responder with forward of the original
message to the list owner.
>most of the packages recognize incorrectly formed requests at least well
>enough to return messages which suggest the correct format.
>From these comments it is obvious that you are not qualified to discuss
ease of use issues.
>No, it was not implemented because they already were honoring the
>long-standing -request convention -- which predates *all* of the
>software packages you mention -- and they saw no reason to go along with
>your attempt to get the world to use your non-standard approach.
LISTSERV honours the long-standing -request convention, and didn't wait
for either you or RFC2142 to do this. However this convention, which
indeed predates mailing list managers, is that the -request address
points to a human, or at least an auto-response explaining how to contact
a human. This should be self-evident considering that the convention
predates mailing list managers. The -server convention is guaranteed to
go to a computer program and not bother a human, which is the whole
point. It is in addition to, not as a replacement for, -request.
>But otherwise, face the fact that you were caught napping by the
>codification of a de facto standard that predates your efforts
LISTSERV is in full compliance with the de facto standard. RFC2142
attempts to redefine this standard to be a totally different thing. This
is why we will not implement it.
>To do otherwise is to thumb your nose at a community's legitimate
>attempt to govern itself, and to declare yourself above that process. It
>will be noted as such, and I daresay you'll be held accountable for it.
Rich, for your information, the mailing list community does not have the
beginning of a clue about what the IETF is or how it works.
Anyway, this discussion is going nowhere and I have work to do.
Eric
