From: Bruce Lilly (blilly@erols.com)
Date: Sun Mar 14 2004 - 16:27:39 CST
Seth Breidbart wrote:
> Expiration is a matter of local policy, which can use whatever it damn
> well pleases, just like killfiles.
>
> Expiration policy can prevent an article from being displayed, just
> like killfiles.
As an incidental side-effect of the article having been removed.
Expiration affects all users and all reader clients on a particular
server. Killfiles affect an individual user using a specific reader
(except for cases where some subset of readers use the same format
killfiles stored in the same place).
> I don't recall anything in our documents limiting expiration policies
> in any way whatsoever, including basing them on the Subject: header.
So local expiration policy could be based on whether an article has
an even (or odd, or prime) number of bytes. So what? None of that
makes any Subject field content part of the protocol. As opposed
e.g. to the Expires field, which does have specific semantics
related to expiration.
> Or are you claiming that the *fact* that *existing software* happens
> to do something should inform our draft?
Not as such.