On Thu, Oct 16, 2003 at 11:57:23AM +0200, Bj?rn Hahnefeld wrote: > Content-Type: text/plain; format=flowed; charset="utf-8"
Ah, so it's not iso-8859-1 after all. Did you recompile sqwebmail with --enable-unicode=utf-8,iso-8859-1 ? If you did, and you want us to help you debug your problem, I suggest you do the following. 1. Deliver a mail into the maildir containing the problem characters. 2. Demonstrate the problem. That is, open the mail in sqwebmail with a browser, produce a screenshot of the browser, and post it to this mailing list. Try it in at least two different browsers on two platforms to prove the problem is not with the browser. Check that they give identical results. 3. Locate the mail message in the Maildir, and *gzip* it # cd /home/foo/Maildir/cur # cat ABCDEF123456:2,S # to check it's the right one # cp ABCDEF123456:2,S /tmp/badmessage # gzip -9 /tmp/badmessage 4. Attach /tmp/badmessage.gz to a message and mail it to this mailing list. (We need to see the complete, exact headers *and* content of the message. Simply forwarding the message using sqwebmail or any other mail client is no good since that will modify the message. Hence taking the file directly out of the Maildir, gzipping it to prevent modification, and then attaching it, is the only acceptable way to look at the content of the message) We will then be able to insert this message into a local Maildir, and see how sqwebmail and our own browsers cope with it; and check whether the message encoding is correct, There are a number of things which might not be right along the chain. My guesses are: (1) you didn't recompile sqwebmail with the right options; or (2) the mail client sent the message wrongly (e.g. it claims charset="utf-8" but in fact the character set used in the message is something else); or (3) your MTA mangled the message before dropping it into the Maildir; or (4) your browser isn't handling utf-8 content correctly. Of course there are other possibilities, such as a bug in sqwebmail, but sqwebmail's multi-lingual support has been around for so long that it seems unlikely. Regards, Brian.
