Cyrus IMAP 3.2.0 released

[ellie timoney]

Hi Marco,

> But it really seems that the LD_PRELOAD or the syslog.so doesn't 
> work for me.

Thanks for including that detail.  Hopefully someone familiar with RedHat can chime in with some insight into the LD_PRELOAD issue!  At a guess, "preventing library injection from intercepting syslog calls" might be a security default of some sort.

> A list of known failing tests for each stable supported version of Cyrus 
> IMAP could be very appreciated, but I understand that this could be 
> difficult to achieve...

Yeah, this would not really be workable, cause even though an individual release of Cyrus doesn't change once it's released, Cassandane changes constantly.  We could maybe publish such a list statically at release time, but it's potentially out of date days later, and it would be a headache to be regularly running Cassandane against all the old builds AND working out which new failures are "known failures" and which are "bugs in Cassandane" etc etc.

I think most people don't run Cassandane very often, so it's easier for them to just reach out for advice when they do.  And if you run Cassandane half a dozen times a day, like I do, you already know which failures can be ignored-for-now and which need investigation.

> My currently failing and now ignored tests are based on failing tests in 
> fedora for 3.0.x, reviewed in my env:
> [...]
>      # This fail on 3.2.0
>      # https://github.com/cyrusimap/cyrus-imapd/issues/2332
>      Caldav.supports_event

This one should be passing on 3.2.0, the issue above has been closed

>      # https://github.com/cyrusimap/cyrus-imapd/issues/2087
>      ImapTest.append-binary
>      ImapTest.fetch-binary-mime
>      ImapTest.urlauth-binary
> 
>      # This one seems to fail randomly on fedora, but in my env always 
> seems to be successful.
>      ImapTest.urlauth2

3.2.0 passes all the imaptest-20190504 tests for me, and I don't skip any.  But I don't see urlauth2 or any of the -binary ones in the output, so maybe upstream imaptest has removed the bad tests.  (The issue quoted was reported for imaptest-20170719)

The rest of these look like about what I'd expect, no surprises. Whew!

Cheers,

ellie