[BUG] yanking base64 encoded e-mail is a fail

[BUG] yanking base64 encoded e-mail is a fail

[Łukasz Stelmach]

Hello.

When I gnus-summary-reply-with-orignal on a message that is single part,
text/plain but base64 encoded (with Content-Transfer-Encoding: base64) I
get a base64 blob as a citation instead of a human readable message.

-- 
Miłego dnia,
Łukasz Stelmach

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

Łukasz Stelmach <lukasz.stelmach@iem.pw.edu.pl> writes:

> When I gnus-summary-reply-with-orignal on a message that is single part,
> text/plain but base64 encoded (with Content-Transfer-Encoding: base64) I
> get a base64 blob as a citation instead of a human readable message.

Yes, I noticed that too, but I can't determine why that's happening.  It
doesn't look like anything has changed in this area, but this is new
behaviour, right?  Anybody?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Yes, I noticed that too, but I can't determine why that's happening.  It
> doesn't look like anything has changed in this area, but this is new
> behaviour, right?  Anybody?

In particular, it's just yanking the stuff from gnus-article-copy, which
contains the undecoded article contents, and isn't very yank-friendly.

But git claims that it's done this since 1997:

deb61c48 lisp/gnus-msg.el     (Lars Magne Ingebrigtsen 1997-03-04 08:29:17 +0000  777)       (let ((message-reply-buffer gnus-article-copy)

How weird.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

Hm.  But this patch was folded into Gnus a few days ago:

+2010-01-17  Chong Yidong  <cyd@stupidchicken.com>
+
+	* message.el (message-mail): Just pass yank-action on to message-setup.
+	(message-setup): Handle (FUN . ARGS) form of yank-action.
+	(message-with-reply-buffer, message-widen-reply)
+	(message-yank-original): Handle non-buffer values of
+	message-reply-buffer (Bug#4080).
+	(message-setup-1): Prefer to save message-reply-buffer as a buffer.

But I can't really see how that would affect anything.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> +2010-01-17  Chong Yidong  <cyd@stupidchicken.com>

D'oh.  I misread the date...  so it's not that...

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

I found the problem.  It's the <#secure thing which made
message-goto-body return the wrong thing.

I've now reverted the change.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Julien Danjou]

[-- Attachment #1: Type: text/plain, Size: 633 bytes --]

On Wed, Nov 24 2010, Lars Magne Ingebrigtsen wrote:

> Yes, I noticed that too, but I can't determine why that's happening.  It
> doesn't look like anything has changed in this area, but this is new
> behaviour, right?  Anybody?

I got this too, and bissect to the last commit actually:

commit 2429ee096e23974cbc3beb2a88998b0e45bdda33
Author: Łukasz Stelmach <lukasz.stelmach@iem.pw.edu.pl>
Date:   Tue Nov 23 21:55:35 2010 +0100

    message-goto-body after <#secure

But I can't explain why. Does a revert of that code fix it for you?

-- 
Julien Danjou
// ᐰ <julien@danjou.info>   http://julien.danjou.info

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: [BUG] yanking base64 encoded e-mail is a fail

[Julien Danjou]

<#secure method=pgpmime mode=sign>
On Wed, Nov 24 2010, Lars Magne Ingebrigtsen wrote:

> I found the problem.  It's the <#secure thing which made
> message-goto-body return the wrong thing.
>
> I've now reverted the change.

I've re-reverted with a fix. Does this seems good to you?

-- 
Julien Danjou
// ᐰ <julien@danjou.info>   http://julien.danjou.info

Re: [BUG] yanking base64 encoded e-mail is a fail

[Sven Joachim]

On 2010-11-24 21:28 +0100, Julien Danjou wrote:

> <#secure method=pgpmime mode=sign>

Hm, this line should not actually have been sent, should it?

> On Wed, Nov 24 2010, Lars Magne Ingebrigtsen wrote:
>
>> I found the problem.  It's the <#secure thing which made
>> message-goto-body return the wrong thing.
>>
>> I've now reverted the change.
>
> I've re-reverted with a fix. Does this seems good to you?

Looks like there's still some flaw if mml tags end up in mails.

Cheers,
       Sven

Re: [BUG] yanking base64 encoded e-mail is a fail

[Lars Magne Ingebrigtsen]

Julien Danjou <julien@danjou.info> writes:

> I've re-reverted with a fix. Does this seems good to you?

It should really return the point where the body starts, which it does
now, but, on the other hand, if you have a <#secure thing in there,
it'll return a different value from where it's putting point, so it
seems inconsistent...  I'm not sure what the <#secure thing is trying to
fix, but I think it sounds like that is being done in the wrong place,
perhaps. 

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [BUG] yanking base64 encoded e-mail is a fail

[Łukasz Stelmach]

<#secure method=pgpmime mode=sign>
Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Julien Danjou <julien@danjou.info> writes:
>
>> I've re-reverted with a fix. Does this seems good to you?
>
> It should really return the point where the body starts, which it does
> now, but, on the other hand, if you have a <#secure thing in there,
> it'll return a different value from where it's putting point, so it
> seems inconsistent...

IMHO if it makes any difference whether message-goto-body returns point
right after mail-header-separator or after <#secure then there is
something wrong. My mistake was not to return the point from the
mssage-goto-body at all. Take a look at <87wro2babo.fsf@kotik.lan>, I
think there is a fix.

> I'm not sure what the <#secure thing is trying to fix, but I think it
> sounds like that is being done in the wrong place, perhaps.

OK, so I made a mess. 

-- 
Miłego dnia,
Łukasz Stelmach