About follup headers on usenet

About follup headers on usenet

[reader]

I seem to recall seeing something in my replay headers when doing a
followup that indicated what I was following up to.

Recently I noticed my header that appears in the composition mode
shows only these headers:
  Newsgroups: comp.unix.shell
  Subject: Re: SED: replace all the underscores [...]
  From: reader@newsguy.com
  Date: Fri, 11 Jan 2008 12:32:44 -0600
  Message-ID: <87myrc43hv.fsf@newsguy.com>
  Organization: Still searching...
  User-Agent: Gnus/5.110007 (No Gnus v0.7) Emacs/22.1 (gnu/linux)
  Gcc: nnml:news.news-0108

Should there be something more?

I do have a setting in .gnus that may effect this:
  (setq message-generate-headers-first t )

Re: About follup headers on usenet

[Reiner Steib]

On Fri, Jan 11 2008, reader@newsguy.com wrote:

> I seem to recall seeing something in my replay headers when doing a
> followup that indicated what I was following up to.

References?  The References header is hidden by default in No Gnus
when composing a message.  It is inserted when sending the message.

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: About follup headers on usenet

[reader]

Reiner Steib <reinersteib+gmane@imap.cc> writes:

> On Fri, Jan 11 2008, reader@newsguy.com wrote:
>
>> I seem to recall seeing something in my replay headers when doing a
>> followup that indicated what I was following up to.
>
> References?  The References header is hidden by default in No Gnus
> when composing a message.  It is inserted when sending the message.

Thanks... has it always been that way?  I mean in last 2yrs or so?

Re: About follup headers on usenet

[reader]

Reiner Steib <reinersteib+gmane@imap.cc> writes:

> On Fri, Jan 11 2008, reader@newsguy.com wrote:
>
>> I seem to recall seeing something in my replay headers when doing a
>> followup that indicated what I was following up to.
>
> References?  The References header is hidden by default in No Gnus
> when composing a message.  It is inserted when sending the message.

But it would seem this should make them appear... no?

   (setq message-generate-headers-first t )

If not that, then what mechanism?

Re: About follup headers on usenet

[Reiner Steib]

On Sun, Jan 13 2008, reader@newsguy.com wrote:

> Reiner Steib <reinersteib+gmane@imap.cc> writes:
>> References?  The References header is hidden by default in No Gnus
>> when composing a message.  It is inserted when sending the message.
>
> Thanks... has it always been that way?  I mean in last 2yrs or so?

Since 2004 (but as said before, only in the development version):

,----
| 2004-11-22  Romain Francoise  <romain@orebokech.com>
| 
| 	* message.el (message-header-format-alist): Add `From' in list
| 	so that it can be sorted.
| 	(message-fix-before-sending): Widen and sort headers before
| 	sending.
| 	(message-hide-headers): Use narrowing to hide headers by moving
| 	them to the top of the buffer and narrowing to the region
| 	underneath.
`----

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: About follup headers on usenet

[Dan Nicolaescu]

Reiner Steib <reinersteib+gmane@imap.cc> writes:

  > On Sun, Jan 13 2008, reader@newsguy.com wrote:
  > 
  > > Reiner Steib <reinersteib+gmane@imap.cc> writes:
  > >> References?  The References header is hidden by default in No Gnus
  > >> when composing a message.  It is inserted when sending the message.
  > >
  > > Thanks... has it always been that way?  I mean in last 2yrs or so?
  > 
  > Since 2004 (but as said before, only in the development version):

And this has the bad property that even if you do
(setq message-hidden-headers nil)

and then remove the "References" header, it will still get added to the
message when the message is sent.
Is this considered a good idea?

Re: About follup headers on usenet

[Ted Zlatanov]

On Tue, 15 Jan 2008 23:22:40 -0800 Dan Nicolaescu <dann@ics.uci.edu> wrote: 

DN> Reiner Steib <reinersteib+gmane@imap.cc> writes:
>> On Sun, Jan 13 2008, reader@newsguy.com wrote:
>> 
>> > Reiner Steib <reinersteib+gmane@imap.cc> writes:
>> >> References?  The References header is hidden by default in No Gnus
>> >> when composing a message.  It is inserted when sending the message.
>> >
>> > Thanks... has it always been that way?  I mean in last 2yrs or so?
>> 
>> Since 2004 (but as said before, only in the development version):

DN> And this has the bad property that even if you do
DN> (setq message-hidden-headers nil)

DN> and then remove the "References" header, it will still get added to the
DN> message when the message is sent.
DN> Is this considered a good idea?

I think it's OK, since 99% of users either want it or don't care.  I'd
leave it on by default.

The remaining 1% should have a clean way of removing that header I
guess.  Is there one?  I don't know of it.

Ted

Re: About follup headers on usenet

[Dan Nicolaescu]

Ted Zlatanov <tzz@lifelogs.com> writes:

  > On Tue, 15 Jan 2008 23:22:40 -0800 Dan Nicolaescu <dann@ics.uci.edu> wrote: 
  > 
  > DN> Reiner Steib <reinersteib+gmane@imap.cc> writes:
  > >> On Sun, Jan 13 2008, reader@newsguy.com wrote:
  > >> 
  > >> > Reiner Steib <reinersteib+gmane@imap.cc> writes:
  > >> >> References?  The References header is hidden by default in No Gnus
  > >> >> when composing a message.  It is inserted when sending the message.
  > >> >
  > >> > Thanks... has it always been that way?  I mean in last 2yrs or so?
  > >> 
  > >> Since 2004 (but as said before, only in the development version):
  > 
  > DN> And this has the bad property that even if you do
  > DN> (setq message-hidden-headers nil)
  > 
  > DN> and then remove the "References" header, it will still get added to the
  > DN> message when the message is sent.
  > DN> Is this considered a good idea?
  > 
  > I think it's OK, since 99% of users either want it or don't care.  I'd
  > leave it on by default.
  > 
  > The remaining 1% should have a clean way of removing that header I
  > guess.  Is there one?  I don't know of it.

Probably the code that inserts the header, should not try to insert it
if it was set to be visible, the header is in the buffer. If it is not,
then the user intentionally deleted it.

Editing References; Bug in ` message-generate-headers'? (was: About follup headers on usenet)

[Reiner Steib]

On Wed, Jan 16 2008, Dan Nicolaescu wrote:

> And this has the bad property that even if you do
> (setq message-hidden-headers nil)
>
> and then remove the "References" header, it will still get added to the
> message when the message is sent.
> Is this considered a good idea?

There was some discussion abut this:

,----
| Subject: Header im message buffer
| From: Oliver Scholz <alkibia...@gmx.de>
| Date: Mon, 19 Mar 2007 19:27:50 +0100
| Message-ID: <45fed62a$0$20297$9b4e6d93@newsspool3.arcor-online.net>
| Newsgroups: de.comm.software.gnus
`----

Andreas Seltenreich proposed a patch (see below), but I didn't have
time to test it.  Anyone?

,----[ http://thread.gmane.org/87ejnblqf5.fsf@gate450.dyndns.org ]
| From: Andreas Seltenreich <andreas+ding@gate450.dyndns.org>
| Subject: dysfunctional code in message-generate-headers
| To: ding@gnus.org
| Date: Mon, 26 Mar 2007 22:07:26 +0200
| Message-ID: <87ejnblqf5.fsf@gate450.dyndns.org>
| 
| Hi,
| 
| there's some discussion in de.comm.software.gnus about an unintuitive
| behavior of Message.  Currently, a "References" header that was
| inserted on buffer creation and deleted by the user is inserted again
| when the mail is being sent.  There's also an older bug report about
| this: <news:87lkok9nnw.fsf@jidanni.org>.
| 
| Looking at the code, it seems that message-generate-headers already
| has some provision to avoid repeated insertion.  However, the code
| doesn't cause any effect at the moment, as it only puts /non-optional/
| headers into message-inserted-headers, while the check to avoid
| repeated insertion only applies to /optional/ ones.
| 
| This looks like a bug to me, and actually putting optional headers
| into message-inserted-headers as in the attached patch results in the
| behavior desired in the bug reports.
| 
| I'd like to propose this patch for No Gnus for now, since it
| essentially enables an old feature for the first time.  WDYT?
`----

--8<---------------cut here---------------start------------->8---
--- message.el	15 Mar 2007 22:41:21 -0000	7.202
+++ message.el	26 Mar 2007 00:37:08 -0000
@@ -5580,6 +5580,7 @@
 		      (if formatter
 			  (funcall formatter header value)
 			(insert header-string ": " value))
+		      (push header-string message-inserted-headers)
 		      (goto-char (message-fill-field))
 		      ;; We check whether the value was ended by a
 		      ;; newline.  If not, we insert one.
--8<---------------cut here---------------end--------------->8---

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: About follup headers on usenet

[reader]

Ted Zlatanov <tzz@lifelogs.com> writes:

> I think it's OK, since 99% of users either want it or don't care.  I'd
> leave it on by default.
>
> The remaining 1% should have a clean way of removing that header I
> guess.  Is there one?  I don't know of it.

Can it be made visible during composition?

  (setq message-generate-headers-first t )

Does not do that.

Re: About follup headers on usenet

[Reiner Steib]

On Wed, Jan 16 2008, reader@newsguy.com wrote:

> Can it be made visible during composition?

,----[ `C-h k <menu-bar> <Field> <Show Hidden Headers>' ]
| <menu-bar> <Field> <Show Hidden Headers> runs the command widen
|   which is an interactive built-in function in `C source code'.
| It is bound to <menu-bar> <Field> <Show Hidden Headers>, C-x n w.
| (widen)
| 
| Remove restrictions (narrowing) from current buffer.
| This allows the buffer's full text to be seen and edited.
`----

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Editing References; Bug in ` message-generate-headers'?

[Andreas Seltenreich]

Reiner Steib writes:

> On Wed, Jan 16 2008, Dan Nicolaescu wrote:
>
>> And this has the bad property that even if you do
>> (setq message-hidden-headers nil)
>>
>> and then remove the "References" header, it will still get added to the
>> message when the message is sent.
>> Is this considered a good idea?
>
[...]
> Andreas Seltenreich proposed a patch (see below), but I didn't have
> time to test it.  Anyone?

Just found this fix again while salvaging changes from my obsolete CVS
working directory.  It's been working inconspicuously for three years,
so I've grown a bit more confident it won't cause notable regressions.

Pushing it now.

regards,
andreas