Deeper freeze for Emacs-23 branch soon

Deeper freeze for Emacs-23 branch soon

[Chong Yidong]

If no major problems with the 23.2.93 pretest crop up during the next
week or so, we'll institute a deep freeze for the Emacs-23 branch.

Once this takes effect, commits to that branch should be for regressions
against 23.2, and documentation fixes, only.  All other bugfixes will go
into the trunk.  We'll deal with exceptions, such as crashes, on a
case-by-case basis.

From now until then, I urge everyone to help with the bugfixing for the
23.3 release.  If you see anything that you think should be a release
blocker, please point it out (and set the severity to "serious" in the
bug tracker).

Thanks!

rmail-reply problem [Re: Deeper freeze for Emacs-23 branch soon]

[Kenichi Handa]

In article <87bp2w4cpt.fsf@stupidchicken.com>, Chong Yidong <cyd@stupidchicken.com> writes:

> If no major problems with the 23.2.93 pretest crop up during the next
> week or so, we'll institute a deep freeze for the Emacs-23 branch.

> Once this takes effect, commits to that branch should be for regressions
> against 23.2, and documentation fixes, only.  All other bugfixes will go
> into the trunk.  We'll deal with exceptions, such as crashes, on a
> case-by-case basis.

> >From now until then, I urge everyone to help with the bugfixing for the
> 23.3 release.  If you see anything that you think should be a release
> blocker, please point it out (and set the severity to "serious" in the
> bug tracker).

I've just found that rmail-reply still inserts raw headers
in a reply buffer as something like these:

  From: =?utf-8?B?0JrQvtC90L7QvSDQkNGE0LDQvdCw0YHRjNC10LLQuNGH?= <zecoxut@vneshbank.ru>
  Subject: =?utf-8?B?0JrQu9C10L3QvtCy0YvQuSDRgdC40YDQvtC/Lg==?=

As a result, sending such a message results in
doubly-decoded headers and thus mail clients receiving such
messages don't show the correct strings.

I have not noticed this problem because I've been using a
special hook to workaround it.  :-(

The simplest and safe fix seems to be the attached one.
But, I'm not sure installing it satisfies the criteria of
"deep freeze" (although I strongly recommend it).

Please judge it.

---
Kenichi Handa
handa@m17n.org

2011-02-21  Kenichi Handa  <handa@m17n.org>

	* mail/rmail.el (rmail-start-mail): Decode "encoded-words" of
	header components.

=== modified file 'lisp/mail/rmail.el'
--- lisp/mail/rmail.el	2011-01-15 03:50:47 +0000
+++ lisp/mail/rmail.el	2011-02-21 02:11:29 +0000
@@ -3449,6 +3449,16 @@
 	(setq yank-action (list 'insert-buffer replybuffer)))
     (setq others (cons (cons "cc" cc) others))
     (setq others (cons (cons "in-reply-to" in-reply-to) others))
+    (setq others
+	  (mapcar #'(lambda (elt)
+		      (cons (car elt) (if (stringp (cdr elt))
+					  (rfc2047-decode-string (cdr elt)))))
+		  others))
+    (if (stringp to) (setq to (rfc2047-decode-string to)))
+    (if (stringp in-reply-to)
+	(setq in-reply-to (rfc2047-decode-string in-reply-to)))
+    (if (stringp cc) (setq cc (rfc2047-decode-string cc)))
+    (if (stringp subject) (setq subject (rfc2047-decode-string subject)))
     (if same-window
 	(compose-mail to subject others
 		      noerase nil

Re: rmail-reply problem [Re: Deeper freeze for Emacs-23 branch soon]

[Chong Yidong]

Kenichi Handa <handa@m17n.org> writes:

> I've just found that rmail-reply still inserts raw headers
> in a reply buffer as something like these:
>
>   From: =?utf-8?B?0JrQvtC90L7QvSDQkNGE0LDQvdCw0YHRjNC10LLQuNGH?= <zecoxut@vneshbank.ru>
>   Subject: =?utf-8?B?0JrQu9C10L3QvtCy0YvQuSDRgdC40YDQvtC/Lg==?=
>
> As a result, sending such a message results in
> doubly-decoded headers and thus mail clients receiving such
> messages don't show the correct strings.
>
> I have not noticed this problem because I've been using a
> special hook to workaround it.  :-(
>
> The simplest and safe fix seems to be the attached one.
> But, I'm not sure installing it satisfies the criteria of
> "deep freeze" (although I strongly recommend it).
>
> Please judge it.

Yes, this ought to be fixed for 23.3.  Your patch seems to do the right
thing; please commit it, and thanks for spotting this.

Re: rmail-reply problem [Re: Deeper freeze for Emacs-23 branch soon]

[Kenichi Handa]

In article <87fwrhz4n6.fsf@stupidchicken.com>, Chong Yidong <cyd@stupidchicken.com> writes:

> Kenichi Handa <handa@m17n.org> writes:
> > I've just found that rmail-reply still inserts raw headers
> > in a reply buffer as something like these:
> >
> >   From: =?utf-8?B?0JrQvtC90L7QvSDQkNGE0LDQvdCw0YHRjNC10LLQuNGH?= <zecoxut@vneshbank.ru>
> >   Subject: =?utf-8?B?0JrQu9C10L3QvtCy0YvQuSDRgdC40YDQvtC/Lg==?=
> >
> > As a result, sending such a message results in
> > doubly-decoded headers and thus mail clients receiving such
> > messages don't show the correct strings.
> >
> > I have not noticed this problem because I've been using a
> > special hook to workaround it.  :-(
> >
> > The simplest and safe fix seems to be the attached one.
> > But, I'm not sure installing it satisfies the criteria of
> > "deep freeze" (although I strongly recommend it).
> >
> > Please judge it.

> Yes, this ought to be fixed for 23.3.  Your patch seems to do the right
> thing; please commit it, and thanks for spotting this.

Ok, just done.

---
Kenichi Handa
handa@m17n.org