[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#22754: 25.1.50; Rmail does not follow RFC 2822

From: Marcin Borkowski
Subject: bug#22754: 25.1.50; Rmail does not follow RFC 2822
Date: Mon, 22 Feb 2016 21:13:32 +0100
User-agent: mu4e 0.9.13; emacs

On 2016-02-22, at 19:14, Glenn Morris <address@hidden> wrote:

> Eli Zaretskii wrote:
>>> From: Marcin Borkowski <address@hidden>
>>> Date: Sun, 21 Feb 2016 11:28:27 +0100
>>> Rmail inserts lower-case field names in the header, which goes against
>>> RFC 2822 (pp. 22 and 23, see
>>> https://tools.ietf.org/html/rfc2822#page-22).  It seems that the problem
>>> affects Cc: and In-reply-to: fields.
>> I use Rmail all the time, and these headers are capitalized in my
>> reply messages.  So please provide a recipe starting from "emacs -Q",
>> there could be some other factor at work here.
> I assume it was a theoretical analysis...
> With sendmail-user-agent, mail-setup inserts "Cc" and "In-reply-to"
> directly (apparently the latter should be "In-Reply-To"?), ignoring the
> case passed via other-headers.
> With message-user-agent, message-mail explicitly uses capitalize on all
> headers.
> I suppose there's no guarantee that some other mail-user-agent won't use
> the headers as supplied, but perhaps that would be a bug in the
> user-agent?

Glenn, you mean what I did was theoretical analysis or what Eli did...?

For me, it was a real problem: I use mu4e as my email client, and it
didn't do the capitalization.  It is quite probable that it's mu4e's
bug, since indeed I could not reproduce it on emacs -Q (without mu4e).
It seems that your email saved me some analysis - you're right,
`message-mail' does the capitalization indeed.  When I have time
(probably Wednesday or Thursday), I'll propose a patch to the author of
mu4e.  For now, I'm closing this bug.

Nevertheless, I still think it is a bit sloppy: it looks like Emacs
depends on the incidental fact that all header field names are
capitalized, but AFAIU this is not a rule stated explicitly in RFC2822,
but just (it seems) a convention they have adopted.  IOW, in the
(unlikely of course - now /this/ is theoretical analysis!) case when
RFC2822 is superseded by another one, not adhering to this convention,
Emacs users will be in trouble, or, as a (not very elegant;-)) Polish
saying goes, "va qnex ohgg" (rot13'd since possibly NSFW). ;-)

Thanks and best regards

Marcin Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

reply via email to

[Prev in Thread] Current Thread [Next in Thread]