Replying from an nnir group

Replying from an nnir group

[Didier Verna]

  Hello,

I access my emails via an nnimap server. I frequently search for old
mails with GG on a group, and then type a reply. Unfortunately, sending
the reply doesn't work. I get a

"no such function: nnir-request-create-group" error. The GCC header is

set to something like that:

"nnir:((query . "blabla") (unique-id . "muxmwtqh4vf.fsf"))"

What I do then is to save the draft, exit the nnir group, get back to
the draft group and type Ds. This is rather inconvenient, and the GCC
header is not set to what I want (i.e. the group from which the original
mail was found).

What would make sense to me is to have the GCC header set to the
original nnimap group, not the nnir one. Is that possible? I have just
looked back at my gnus-message-archive-* config, but it seems to date
back from the days I was doing nnml, and it doesn't seem to apply to
IMAP servers.

Any ideas?

Thanks!

-- 
Resistance is futile. You will be jazzimilated.

Scientific site:   http://www.lrde.epita.fr/~didier
Music (Jazz) site: http://www.didierverna.com

Re: Replying from an nnir group

[Dave Goldberg]

> I access my emails via an nnimap server. I frequently search for old
> mails with GG on a group, and then type a reply. Unfortunately, sending
> the reply doesn't work. I get a

> "no such function: nnir-request-create-group" error. The GCC header is

> set to something like that:

> "nnir:((query . "blabla") (unique-id . "muxmwtqh4vf.fsf"))"

> What I do then is to save the draft, exit the nnir group, get back to
> the draft group and type Ds. This is rather inconvenient, and the GCC
> header is not set to what I want (i.e. the group from which the original
> mail was found).

> What would make sense to me is to have the GCC header set to the
> original nnimap group, not the nnir one. Is that possible? I have just
> looked back at my gnus-message-archive-* config, but it seems to date
> back from the days I was doing nnml, and it doesn't seem to apply to
> IMAP servers.

Probably the setting of gcc-self to t.  A posting style that recognizes you're replying from an nnir group and setting Gcc appropriately would be the most likely solution.  There is a macro called nnir-article-group that might be useful for getting the original group name.

-- 
Dave Goldberg
david.goldberg6@verizon.net

Re: Replying from an nnir group

[Andrew Cohen]

>>>>> "dvl" == Didier Verna <didier@xemacs.org> writes:

    dvl>   Hello, I access my emails via an nnimap server. I frequently
    dvl> search for old mails with GG on a group, and then type a
    dvl> reply. Unfortunately, sending the reply doesn't work. I get a

    dvl> "no such function: nnir-request-create-group" error. The GCC
    dvl> header is

    dvl> set to something like that:

    dvl> "nnir:((query . "blabla") (unique-id . "muxmwtqh4vf.fsf"))"

    dvl> What I do then is to save the draft, exit the nnir group, get
    dvl> back to the draft group and type Ds. This is rather
    dvl> inconvenient, and the GCC header is not set to what I want
    dvl> (i.e. the group from which the original mail was found).

    dvl> What would make sense to me is to have the GCC header set to
    dvl> the original nnimap group, not the nnir one. Is that possible?
    dvl> I have just looked back at my gnus-message-archive-* config,
    dvl> but it seems to date back from the days I was doing nnml, and
    dvl> it doesn't seem to apply to IMAP servers.

The simplest solution is to use "A W" before you reply---this will take
you to the original group for the article.

Arguably the reply function should do this for you. gnus-warp-to-article
is a no-op for real articles so it shouldn't hurt to just add this at
the beginning of all the reply functions. And I guess forwarding
functions too. I'll have to think about this...

Re: Replying from an nnir group

[Steinar Bang]

>>>>> Didier Verna <didier@xemacs.org>:

> I access my emails via an nnimap server. I frequently search for old
> mails with GG on a group, and then type a reply. Unfortunately, sending
> the reply doesn't work. I get a

> "no such function: nnir-request-create-group" error. The GCC header is

> set to something like that:

I always to an `A W' on the article before replying.  Then I move to
that article in the orginal nnimap group.  All sender setup variables
are set correctly, and the message is marked as replied to.

(It could be that this is no longer necessary with Andy Cohen's new
changes...?  I haven't pulled them yet.  Will do so now)

Re: Replying from an nnir group

[Didier Verna]

david.goldberg6@verizon.net (Dave Goldberg) wrote:

> Probably the setting of gcc-self to t.  A posting style that
> recognizes you're replying from an nnir group and setting Gcc
> appropriately would be the most likely solution.  There is a macro
> called nnir-article-group that might be useful for getting the
> original group name.

  So, I got a solution:

gnus-message-archive-method:

(t dummy)

gnus-message-archive-group:

((cond ((message-news-p)
        nnimap+gmail:sent-news)
       ((stringp gnus-newsgroup-name)
        (cond ((zerop (length gnus-newsgroup-name))
               nnimap+gmail:sent-mails)
              ((string-match ^nnir: gnus-newsgroup-name)
               (aref (aref nnir-artlist 0) 0))
              (t gnus-newsgroup-name)))
       (t nnimap+gmail:sent-mails)))


Of course, the nnir part is dirty and probably fragile. It will work
only when doing searches on a single group, but that's a start. In fact,
I still don't know how to figure out the exact article (which one in
nnir-artlist) I'm replying to. Maybe an nnir expert knows better.

-- 
Resistance is futile. You will be jazzimilated.

Scientific site:   http://www.lrde.epita.fr/~didier
Music (Jazz) site: http://www.didierverna.com

Re: Replying from an nnir group

[Didier Verna]

Andrew Cohen <cohen@bu.edu> wrote:

> The simplest solution is to use "A W" before you reply---this will
> take you to the original group for the article.

  Hmmm, I didn't know this command, thanks.

-- 
Resistance is futile. You will be jazzimilated.

Scientific site:   http://www.lrde.epita.fr/~didier
Music (Jazz) site: http://www.didierverna.com

Re: Replying from an nnir group

[Didier Verna]

I wrote:

> ((cond ((message-news-p)
>         nnimap+gmail:sent-news)
>        ((stringp gnus-newsgroup-name)
>         (cond ((zerop (length gnus-newsgroup-name))
>                nnimap+gmail:sent-mails)
>               ((string-match ^nnir: gnus-newsgroup-name)
>                (aref (aref nnir-artlist 0) 0))
>               (t gnus-newsgroup-name)))
>        (t nnimap+gmail:sent-mails)))
>
>
> Of course, the nnir part is dirty and probably fragile. It will work
> only when doing searches on a single group, but that's a start. In
> fact, I still don't know how to figure out the exact article (which
> one in nnir-artlist) I'm replying to. Maybe an nnir expert knows
> better.

  And in light of the recent replies, I guess I can do

(nnir-article-group (gnus-summary-article-number))

instead, and get a cleaner solution.

-- 
Resistance is futile. You will be jazzimilated.

Scientific site:   http://www.lrde.epita.fr/~didier
Music (Jazz) site: http://www.didierverna.com

Re: Replying from an nnir group

[Andrew Cohen]

>>>>> "dvl" == Didier Verna <didier@xemacs.org> writes:

    dvl>   Hello, I access my emails via an nnimap server. I frequently
    dvl> search for old mails with GG on a group, and then type a
    dvl> reply. Unfortunately, sending the reply doesn't work. I get a


[...]


    dvl> What would make sense to me is to have the GCC header set to
    dvl> the original nnimap group, not the nnir one. Is that possible?


OK, I've just pushed a fix that should properly set both the posting
style and the gcc when mailing/replying from an nnir summary
buffer. Please test and let me know if it works (and hopefully doesn't
screw up these things for non-nnir summaries).

Regards,
Andy

Re: Replying from an nnir group

[Didier Verna]

Andrew Cohen <cohen@bu.edu> wrote:

> OK, I've just pushed a fix that should properly set both the posting
> style and the gcc when mailing/replying from an nnir summary
> buffer. Please test and let me know if it works (and hopefully doesn't
> screw up these things for non-nnir summaries).

  Thanks, it seems to be working well so far.

-- 
Resistance is futile. You will be jazzimilated.

Scientific site:   http://www.lrde.epita.fr/~didier
Music (Jazz) site: http://www.didierverna.com