expiring mail?

expiring mail?

[Nick Papadonis]

Hi.

I am trying to expire some of my megabytes of mail.  I have two
questions:

1) When doing a C-U in the group buffer, why does gnus say I have SO
MANY messages in a group?  I know there arn't close to the amount
stated.

2) Is there a simple way to expire all articles marked 'O'?  Like a
function to call inside the summary buffer?  I guess total expire
doesn't work on 'O'-old articles?

Any insight appreciated. Thanks!

-- 
- Nick

Re: expiring mail?

[Kai Großjohann]

On 21 Mar 2001, Nick Papadonis wrote:

> 1) When doing a C-U in the group buffer, why does gnus say I have SO
> MANY messages in a group?  I know there arn't close to the amount
> stated.

Holes in article number sequence.  You can compress the sequence by
moving all articles from a group into the same group.

Gnus uses an estimate based on first and last article number because
of efficiency: this data can be had from an NNTP server very quickly,
and the exact number is hard to get.

> 2) Is there a simple way to expire all articles marked 'O'?  Like a
> function to call inside the summary buffer?  I guess total expire
> doesn't work on 'O'-old articles?

You can use the limiting commands (start with `/') to view exactly the
articles you want to expire (mark as expirable?).  Then you can use `M
P a' to mark all these articles as processable, then `E' or `M-& E' to
mark them expirable.

total-expire does work on `O' articles because they are read.  See the
node on read vs unread articles in the info file.

kai
-- 
Be indiscrete.  Do it continuously.

Re: expiring mail?

[Paul Jarc]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> Gnus uses an estimate based on first and last article number because
> of efficiency: this data can be had from an NNTP server very quickly,
> and the exact number is hard to get.

WIBNI we could configure Gnus to use the article count from
-request-group instead?  Say, by a group/topic parameter?  Most of the
time, I'd prefer correctness to efficiency.


paul

Re: expiring mail?

[Nick Papadonis]

prj@po.cwru.edu (Paul Jarc) writes:

> Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> > Gnus uses an estimate based on first and last article number because
> > of efficiency: this data can be had from an NNTP server very quickly,
> > and the exact number is hard to get.
> 
> WIBNI we could configure Gnus to use the article count from
> -request-group instead?  Say, by a group/topic parameter?  Most of the
> time, I'd prefer correctness to efficiency.
> 
> 
> paul
> 
I agree.  Even some way of getting this number reset.  IMHO copying
messages in and out of a group seems like a hack.

-- 
- Nick

Re: expiring mail?

[Paul Jarc]

Nick Papadonis <npapadon@yahoo.com> writes:
> prj@po.cwru.edu (Paul Jarc) writes:
> > WIBNI we could configure Gnus to use the article count from
> > -request-group instead?  Say, by a group/topic parameter?  Most of the
> > time, I'd prefer correctness to efficiency.
> 
> I agree.  Even some way of getting this number reset.  IMHO copying
> messages in and out of a group seems like a hack.

That won't be necessary with the new version of nnmaildir.  Article
numbers are dynamically assigned each time you open the server, so the
only holes will be the result of articles that were expired during
that same Gnus session.


paul