From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3)
Description of problem:
Macintosh Eudora clients fail when connecting via POP to dovecot on
FC3. A possible explanation is at:
"Could be that it's trying to login using AUTH PLAIN command, which
Dovecot 0.99 doesn't really support correctly with POP3. 1.0-tests have
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. run dovecot on FC3 server
2. try to connect via POP from Eudora on a Mac
Actual Results: "Unsupported authentication mechanism" message.
IMAP works fine, other POP clients work fine.
Timo: On the presumption that this is fixed in 1.0 I just did a diff
between 0.99-11 and 1.0-test59 and looked for diffs in the pop3-login
auth code. Hmm... quite a few. Do you happen to recall what the
specific fix/issue was or is just a case where in a largely new body
of code things just seem to work now that didn't before?
I don't have an easy way to test this anyway, I'd have to locate a
Mac (doable) and configure things. I'm also wondering if this
shouldn't just wait for the official 1.0 release. Do you anticipate
that being soon?
Fixed in 0.99.12:
- POP3: Don't return PLAIN in SASL list. We don't support initial SASL
responses, so it only breaks with most clients that try to use it.
If the fix is in 0.99.12 - and presumably not a huge change - is there
any chance of an update? I have a server deployment which needs this.
(I'm surprised more people haven't been bitten by this - Eudora is
still pretty popular on Macs.)
This bug also affects RHEL4 and no update has been released for it. I
had to build the rawhide FC4 dovecot which fixes the problem and
hasn't appeared to break anything else yet.
Note, for RHEL4 the fix is contained in errata RHBA-2005:237 which as far as I
can tell is scheduled for RHEL4-U2
A fix for FC3 was released in fc3-updates, dovecot-0.99.13-3.FC3 released on
2005-01-17, closing ...
My RHEL4-U3 install still shows dovecot-0.99.11-2.EL4.1 as the most recent dovecot version, we're still
hitting this bug. Can't find any mention of RHBA-2005-237. What happened?