Re: [OT] Newsgroups Support for "Cancel", "Supersede", etc.
- Subject: Re: [OT] Newsgroups Support for "Cancel", "Supersede", etc.
- From: Richard Kettlewell <rjk@greenend.org.uk>
- Date: Mon, 16 May 2005 14:16:57 GMT
- Cancel-lock: sha1:d5xJkpH9BCmkXNI4HLLzNbbY+3o=
- Mail-copies-to: nobody
- Newsgroups: uk.net
- Organization: Personal Nuclear Deterrents, Inc.
- References: <d69av1$1v9g$1@godfrey.mcc.ac.uk> <slrnd8h2ri.sdb.chris@ccserver.keris.net>
- User-agent: Gnus/5.1007 (Gnus v5.10.7) Emacs/21.4 (gnu/linux)
- Xref: news.mcc.ac.uk uk.net:81643
Chris Croughton <chris@keristor.net> writes:
> There are a number of reasons "Supersede" and "Cancel" don't work
> universally:
>
> Some servers refuse to action them because of abuse (forged messages
> trying to usurp a poster's messages).
It's a shame cancel-lock or something similar doesn't seem to have
really taken off.
> Propagation delays (what if a '"Supersede" or "Cancel" arives before
> the message it is supposed to delete? The message may have been
> deleted 'upstream' so it can't keep the information around forever).
It only really needs to keep them as long as, and in the same database
as, the message ID would be kept had the article arrived and not been
cancelled (however long that may be at that site).
> Few clients honour either (many don't see cancels at all), so if the
> user gets the message into their local newsbase before the server has
> deleted the original they will still see the original.
Supersedes would be easy enough for such a client to honor. Cancels
would be more inconvenient, though not necessarily impossible.
--
http://www.greenend.org.uk/rjk/
|
|
|
