bugtracker sets bad Mail-Followup-To

bugtracker sets bad Mail-Followup-To

[Sven Joachim]

The bugtracker sets not only Reply-To, but also Mail-Followup-To to the
submitter and nnnn@debbugs.gnu.org.  It also does this when replying to
an existing report, which is very bad because it cuts off previously
CC'ed people from Reply-all followups.

Sven

Re: bugtracker sets bad Mail-Followup-To

[Glenn Morris]

Sven Joachim wrote:

> The bugtracker sets not only Reply-To, but also Mail-Followup-To to the
> submitter and nnnn@debbugs.gnu.org.  It also does this when replying to
> an existing report, which is very bad because it cuts off previously
> CC'ed people from Reply-all followups.

It's an attempt to reduce the massive mail duplication that existed
before, when people kept replying to nnn@ and bug-gnu-emacs.

Is the right thing to do:

1) When I reply to nnn@, the mail that debbugs sends out should have
neither MFT nor RT in the headers. Then people can just followup as
normal.

2) When I send a new mail to bug-gnu-emacs or submit@, the new report
that goes out should either:

a) Have MFT set to nnn@, and no RT header (same as at present, bug
remove RT).

b) Have To: set to nnn@ rather than bug-gnu-emacs. (I think this is
preferable?)


If so, I will _try_ to implement it.

RE: bugtracker sets bad Mail-Followup-To

[Drew Adams]

Is this perhaps related to the fact that (at least some) bug reports sent to
emacs-pretest-bug@gnu.org never appear at bug-gnu-emacs@gnu.org?

I get an Ack message from bug-gnu-emacs@gnu.org, but the bug report itself never
gets sent on to bug-gnu-emacs@gnu.org. See thread "bug tracker down again?".
(FWIW, my mail client is Outlook 2003.)

RE: bugtracker sets bad Mail-Followup-To

[Glenn Morris]

Drew Adams wrote (on Fri, 8 Jan 2010 at 10:14 -0800):

> Is this perhaps related to the fact that (at least some) bug reports
> sent to emacs-pretest-bug@gnu.org never appear at
> bug-gnu-emacs@gnu.org?

No.

> I get an Ack message from bug-gnu-emacs@gnu.org, but the bug report
> itself never gets sent on to bug-gnu-emacs@gnu.org. See thread "bug
> tracker down again?".

From past experience, I _think_ the bug-gnu-emacs mailing list
silently discards mails above a certain size.

RE: bugtracker sets bad Mail-Followup-To

[Drew Adams]

> > I get an Ack message from bug-gnu-emacs@gnu.org, but the bug report
> > itself never gets sent on to bug-gnu-emacs@gnu.org. See thread "bug
> > tracker down again?".
> 
> From past experience, I _think_ the bug-gnu-emacs mailing list
> silently discards mails above a certain size.

The bug report I mentioned in the thread is #5325.
It has only 1200 chars, and no attachments.

bug#5325 [was RE: bugtracker sets bad Mail-Followup-To]

[Glenn Morris]

Drew Adams wrote (on Fri, 8 Jan 2010 at 10:55 -0800):

> The bug report I mentioned in the thread is #5325.

I think there was some transient debbugs weirdness going on at the
time you sent that. There is no need to send the bug again, I'm sure
we're all aware of it now.

Re: bugtracker sets bad Mail-Followup-To

[Stefan Monnier]

> It's an attempt to reduce the massive mail duplication that existed
> before, when people kept replying to nnn@ and bug-gnu-emacs.

> Is the right thing to do:

> 1) When I reply to nnn@, the mail that debbugs sends out should have
> neither MFT nor RT in the headers. Then people can just followup as
> normal.

> 2) When I send a new mail to bug-gnu-emacs or submit@, the new report
> that goes out should either:

> a) Have MFT set to nnn@, and no RT header (same as at present, bug
> remove RT).

> b) Have To: set to nnn@ rather than bug-gnu-emacs. (I think this is
> preferable?)

> If so, I will _try_ to implement it.

IMO, the "right" thing to do is to replace bug-gnu-emacs@gnu.org and
emacs-pretest-bug@gnu.org by nnn@debbugs.gnu.org everywhere in the
headers.


        Stefan

Re: bugtracker sets bad Mail-Followup-To

[Glenn Morris]

Stefan Monnier wrote:

> IMO, the "right" thing to do is to replace bug-gnu-emacs@gnu.org and
> emacs-pretest-bug@gnu.org by nnn@... everywhere in the headers.

A version of this is now in place, for new reports only.
As a free bonus, X-Debbugs-CC is now handled properly (ie converted to
a real Cc in the outgoing message).