On 21-02-14 15:05, Peter Milesson wrote:
>
>
> On 2014-02-21 14:05, Paul J Stevens wrote:
>> On 21-02-14 13:41, Peter Milesson wrote:
>>
>>> In zipped attachment, I send the result from running the testfetch.txt
>>> through nc, as you described above. The result when running through
>>> 3.1.10 is just plain wrong, and the result running through 3.0.2 is
>>> correct. The ONLY difference is DBMail version. All libraries are
>>> identical, the data set is identical, everything is compiled a few
>>> minutes apart in the same machine. But DBMail 3.1.10 is just plain
>>> wrong.
>>
>>
>> I don't see an attachment.
>>
> I'm quite sure I attached it, but here we go again...
How am I supposed to compare here? Both files contain results for
different mailboxes...
Also, I don't understand why you're saying the 3.1.10 results are 'plain
wrong'. Looks ok to me, at least from an IMAP perspective.
If the user/mailboxes opened are indeed identical, then I want a full
debug trace of:
x login user password
x select mailbox
x logout
No *fetch* is required.
Thing is: the number of messages *and* the uidvalidity values are
completely different between the logs. This is clear proof of the fact
that *different* mailboxes are actually accessed.
I can think of only one explanation for that - assuming the dataset and
imap commands are indeed identical: usermap. You can verify this by:
SELECT * FROM dbmail_mailboxes WHERE mailbox_idnr IN (129, 174);
since those are the UIDVALIDITY values reported. See if those point to
different owner_idnr values.
If that doesn't help; send me the full trace mentioned above.
--
________________________________________________________________
Paul J Stevens pjstevns @ gmail, twitter, github, linkedin
www.nfg.nl/[email protected]/+31.85.877.99.97
_______________________________________________
DBmail mailing list
[email protected]
http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail
