[no subject]

["D. J. Bernstein" <djb@xxxxxxxx>]

A straw poll on the ietf-submit mailing list in May 1997 revealed
widespread disgust for Change-History. Supporters are clearly in the
minority on the mailing list. This document should not be in Last Call.

> Time is of the essence.

Tim Goodwin had a document more than a year ago describing what people
actually do in practice. In short, the problem has already been solved
by bilateral agreement. A well-known port is helpful but not essential.

> The current comments are not timely,

Bullshit. The current objections have been raised repeatedly over the
past two years. Sample comments from 1997 on the Change-History garbage:

   * Ned Freed (PMDF): ``an unimplementable solution to a non-problem''
   * Philip Hazel (Exim): ``don't like that''
   * me (qmail): ``counterproductive''
   * Christian Kuhtz: ``bogus''

There were more detailed objections; Gellens ignored all of them. I also
asked, for example, why the allowable cleanups were restricted:

   ``Is it really necessary to micromanage the set of possible fixups
   according to one group's current inexperience, prejudice, and lack of
   imagination? Why can't sysadmins and users make whatever fixups they
   want?''

Again there were more detailed objections. Gellens ignored all of them.

> nor do they contain specific suggestions for changes to the document.

Many specific suggestions appear in the ietf-submit archive. Gellens
ignored all of them.

You're deluding yourself if you think that users will accept the Gellens
requirements. What he calls ``dangerous modifications'' are what users
call ``helpful defaults.''

---Dan
Smaller, faster, safer than inetd+tcpd. http://pobox.com/~djb/ucspi-tcp.html