On 12-09-26 20:59:51, Christian Ebert wrote: > * Ambrevar on Wednesday, September 26, 2012 at 21:43:18 +0200 > > [Using Mutt-1.5.21 with FreeBSD and Arch Linux] > > > > Hi there, > > > > I noticed a nasty behaviour I do not understand. It happens if the e-mail > > contains latin characters with acute, like 'é', but no unicode character not > > covered by latin1. The text editor -- Emacs, but I tested with others too -- > > will correctly set it to UTF-8. > > > > file /tmp/mutt-... > > > > will confirm this. > > > > When I close the editor, thus swithing back to Mutt, it sees the content as > > iso8859-1. This is on the "Mutt Compose" screen, right before actually > > sending > > the mail. Of course I can convert it at this very same point with the > > 'edit-type' function. But I do not want to do it manually. It would be nice > > if I > > could send all my mails in UTF-8 automatically. > > > > If I add a unicode character not covered by Latin1 (e.g. '€'), then the > > e-mail > > is correctly marked as being utf-8. > > > > All my locale variables are set to "en_US.UTF-8". And > > > > :set &charset ?charset > > > > returns 'utf-8'. > > > > So I guess it comes from Mutt that fails at guessing the proper character > > encoding. I do not know how it works internally, but I do not have so much > > time > > right now to crawl in the source code. > > > > To reproduce the issue: write a e-mail with the sole letter 'é' as content. > > > > Note that the e-mail is not screwed up. It gets properly converted from > > utf-8 to > > latin1, 'é' goes from a9c3 (utf-8) to e9 (latin1). > > > > Any clue? > > man 5 muttrc: > > send_charset > Type: string > Default: "us-ascii:iso-8859-1:utf-8" > > A colon-delimited list of character sets for outgoing messages. Mutt > will > use the first char- acter set into which the text can be converted > exactly. If your $charset is not "iso-8859-1" and recipients may > not > understand "UTF-8", it is advisable to include in the list an appropri- > ate widely used standard character set (such as "iso-8859-2", > "koi8-r" > or "iso-2022-jp") either instead of or after "iso-8859-1". > > In case the text cannot be converted into one of these exactly, mutt > uses > $charset as a fall- back. > > I have: > set send_charset="us-ascii:iso-8859-1:iso-8859-15:windows-1252:utf-8" > > And when I write €, this email will probably be iso-8859-15. > > It's all about saving old-fashionedly some bits over the wire. > > -- > theatre - books - texts - movies > Black Trash Productions at home: http://www.blacktrash.org > Black Trash Productions on Facebook: > http://www.facebook.com/blacktrashproductions
Thank you so much for this quick answer! Well, it definitely works! I wasn't aware of the send_charset variable: I must have missed it in the not-so-short man page and in the wiki. The thing is, I was looking for 'encoding', not 'charset'. :/ Saving bits may be understandable, but in my humble opinion universality and simplicity is more important. UTF-8 saves us a lot of pain, that's why I'm a UTF-8 advocate -- and for a lot of other reasons too. Cheers.