Re: Archive (was: Issues outstanding)

["Charles Lindsey" <chl@xxxxxxxxxxxxxxxx>]

In <41CB3EC5.2E92@xxxxxxxxxxxxxxxxx> Frank Ellermann <nobody@xxxxxxxxxxxxxxxxx> writes:

>Charles Lindsey wrote:

>  [3. Mail-Copies-To and Posted-And Mailed}
>>> I just don't want to spend another few weeks debating the issue.

>> A quick head count, perhaps?

>Count me against Posted-And-Mailed.  You say that some UAs know
>it, that's fine, but my UA does not, and in the extremely rare
>cases when I get a "courtesy copy" I want this info in the body
>of the message.  Otherwise I would love to auto-killfile any
>unsolicited Posted-And-Mailed without info in the body, but my
>UA also has no killfile (for news).  This header is a PITA and
>conflicts with any minimal netiquette I know of.

All a UA can do with that header is to display it (and a decent UA should
be configurable to display and header the user asks for, though sadly that
is not always the case). But there is no way we could mandate putting
stuff in the body. But I think Mail-Copies-To is the header we really need
to decide upon; Posted-And-Mailed just comes with it as an added extra.

>  [6. Remove filename parameter from the Archive header]
>> Please do bear in mind that it is hard to add new stuff to
>> existing headers in the future

>The complete "Archive:" is dead on arrival, even without the
>filename.  Some existing software supports X-NoArchive somehow,
>e.g. Google now interprets it as "expires after 7 days".

Again, a decent UA will display it if a user asks. But normal UAs don't
need to see it - it is just specialized archiving agents that need to see
and act upon it, and Google is the obvious example of such an agent.

>  Maybe a recommendation in USEAGE about a configurable
>"Expires:" default would be better than X-No or Archive: NO ?

No, because there are legitimate uses of "Expires:" (keeping FAQs around
until the next issue arrives) that are quite consistent with "Archive:
yes".

>> That is the whole beauty of MIME-style parameters,

>After 2231 nothing about MIME parameters is still beautiful :-(

>And the filename stuff just makes no sense for different file
>systems, let alone I18N.  Martin's Archived-At: is the future.

Martin's Archived-At: fulfils a totally different function. It is for
where the author has already arranged for a permanent archive, not for
where is is going to be achived by a third party. Though, come to think of
it, it might be OK for FAQs in news.answers where the author knows in
advance that it will automatically get archived by rtfm.mit.edu and its
mirrors.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl@xxxxxxxxxxxxxxxx      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5