Re: Re: Re: Differences between RFC 2822 and Usefor

From: John Stanley (stanley@peak.org)
Date: Fri May 02 2003 - 20:13:19 CDT


Charles Lindsey (chl@clerew.man.ac.uk):

>Actually, the really important thing to say is that you MUST NOT say
>"Re: Re: ...".

Now it is "really important" to enforce some structure on an unstructured
header. Why is it 'really important', given that it still hasn't been
proven that is important at all?

>But OTOH you don't want to put implementors
>to have to bother with "Re: Re: Differences between RFC 2822 and Usefor",
>or "Re[2]: ..." or other horrors (because current practice seems to be
>that you only ever write one "Re: ").

Implementors don't HAVE to bother with anything in the Subject header, the
threading of an article is defined by the REFERENCES header, not the
Subject header. And no, we are NOT suddenly creating this REFERENCES
header, it has been MANDATORY in followups for what, ten years now? No,
I'm sorry, that's more than FIFTEEN years. (RFC1036 came out in 1987. It's
now 2003, in case you missed it. In December of this year, it will be
SIXTEEN years. By the time USEFOR is done, probably twenty.)

If implementors haven't gotten References down pat by now, they don't
intend on doing it, and if they don't intend on following the standards I
see absolutely NO reason for this group to continue a kludge for them.

> But IMAP is not my problem, at the moment.

Nor is the failure of an implementor to implement a header that has been
defined for more than 15 years our problem.

> However, the time to discuss it is
> when we come to do the USEAGE ...

The hell with proper spelling if Charles' cross-referencer cannot handle
correctly spelled words, huh? I was going to suggest that we call the
minimalized versions of USEFOR "USELESS", but I guess we'll have to call
them "USELES" so the cross-referencer won't have a problem.

I am waiting with breathless anticipation the reports of computer systems
around the planet crashing and burning just because I started the subject
of this message with "Re: Re: Re:". If two is bad, three must be badder,
yes?




This archive was generated by hypermail 2.1.7.