On Fri, 4 Dec 1992 12:05:07 EST "(Wayne Smith)" <[log in to unmask]>
said:
> Wagner is telling IBM that repair of SMTP is a waste of time? Can you
> tell me some of the story behind your remark, please? (Is SMTP being
> rewritten?)
Most of the story is available from the IBMTCP-L logs. It happened around
february of this year. SMTP is not being rewritten. The code in SMTP that
folds headers longer than 80 is buggy and this was preventing all
Lithuanian users from accessing BITNET (it worked only in one direction,
so no mailing list subscription or the like was possible). I wanted IBM
to fix it of course. Wagner claimed first that SMTP was not violating any
standard and thus IBM was free not to fix it, and then after this
statement was proved wrong by a RFC expert he said this was a waste of
manpower given the fact that it will not completely solve the problem of
long userids (only most of them) and that the VM mailer would support
Netdata format in the near future.
IBM decided not to fix the problem. This was a serious blow for the
credibility of BITNET in Nordic countries, as Lithuania was unable to
access BITNET for several months (they then got .LT registered and the
problem was bypassed because the addresses became shorter). Finland has
already started phasing out BITNET, with 6 nodes out of 26 deleted this
month; FINHUTC, the home of one of the largest LISTSERV archives in the
network, if not THE largest, and a pioneer in the development of BITNET
and LISTSERV, is going away next month or in february. We still don't
know where we will put the 92 public lists it was hosting. My position
was removed from the organizational charts last week, along with a
decision to downgrade the NJE service from a "service" to an
"application", like TELNET, which requires no particular coordination.
For reasons which would be long to explain that doesn't affect my job in
the immediate future. However, BITNET is now definitely on its way out.
Note that these decisions were made *after* I wrote LMail (I was informed
the day after the original announcement). So this is not the reason I
wrote it; in fact, the news made it look like I had worked in vain.
I'm not blaming any of this on Wagner and IBM, it was only one part of
the picture and I am fully aware of that. But the fact that, without
LMail, BITNET would still be unable to handle these addresses and, as
seen by external decision-makers, "BITNET" not only did not do anything
to fix a problem preventing an entire country from being served but
actually asked IBM not to fix it because it isn't worth it really did not
give the kind of impression that might have prevented this move. And I'm
cc:-ing Wagner on this so he gets a chance to flame me.
Eric
|