Hi.
We upgraded our company's Zimbra Community Edition installation a couple of days ago and we've been having some troubles ever since with the headers of some incoming mail messages, they are encoded in ISO-8859-15 quoted printable encoding. All the international letters (I say again, quoted printable, this shouldn't be happening) are replaced with a question mark symbol.
We upgraded from 5.0.18 where this was not a problem.
I have only confirmed this in the web client as I do not use an off-line client.
Example (taken from the source of the e-mail by 'Show Original'):
Code:
Subject: Verkbei�ni �thluta� til ��n
The subject should say (in a perfect world):
Code:
Subject: Verkbeiðni úthlutað til þín
Original subject header (from the sender):
Code:
Subject: =?ISO-8859-1?Q?Verkbei=F0ni=20=FAthluta=F0=20til=20=FE=EDn?=
Originally the e-mails were not encoded in this style from the sender (automatic system), they were written in iso-8859-15 encoding with no quoted printables. Zimbra (5.0.18) had a problem with displaying the international characters correctly (converting them from iso-8859-15 to utf-8) so I had to change the source code of the report system so that it sends quoted printable encoded emails instead of plain iso-8859-15.
To make sure that this was not something connected to the upgrade, I set up a new test server (6.0.3) on a different site and sent an e-mail using the same reporting utility to that server. The e-mail was displayed precisely in the same manner as on the production e-mail server.
I have not found any relevant post or bug that can explain this or resolve this.
Do any of you believe that I can change this behavior by some configuration parameters (perhaps Java configuration)?
Has this been an issue for anyone else (and if so, what did you do to resolve it)?
(Possibly related)
If I send an e-mail using no quoted printable encoding on an iso-8859-15 encoded email with international characters (ðæþ for instance), Zimbra Web access displays the international characters in the subject and body as a question mark symbol in the list view, but the quick preview displays the international characters in the body correctly but displays question mark symbols for each international character in the subject.
Bugzilla
Wiki
Source
ZCS 6.0.3 reading incoming mail subject and body incorrectly? 
Linear Mode
