Gary,
The ASF has clear, documented guidelines of what constitutes a release
and what a release tester is supposed to do. I follow those guidelines.
You say you have better ones that take precedence. Uhm, ok.
In general I think I am quite a flexible person. A decent argument can
easily sway my vote. It happened in the past. I don't see one in this
case. I also don't know what else I could add that would be considered
constructive along the lines that the quality of our releases is
important and if it's reasonably possible we should try our best to
ensure said quality. If you can think of such a positive, constructive
argument, please consider that I made it (I tried my best in my own
words). I myself am guilty of being more "flexible" in the past, I don't
think this is a good thing and I will try to correct that.
That said, as I mentioned before, my guidelines also say that a release
cannot be vetoed. So now we (pmc) need to make a choice.
Cheers,
Hadrian
On 01/29/2015 08:33 AM, Gary Tully wrote:
Hadrian,
the true writer really is the reader. If the guidelines replaced "
compile it as provided", with " compile and tests it as provided" your
suggestion would have some merit.
We even document[1] the maven skip tests build option.
[1] http://activemq.apache.org/building.html
On 29 January 2015 at 12:06, Hadrian Zbarcea <[email protected]> wrote:
Gary,
It's a very simple matter. We release source distributions, not binaries.
One cannot reliably build binaries from the source distro. Even with -fae,
everything else depending stomp is skipped, so failures may be hiding other
problems. I would suggest carefully reviewing the ASF release policy [1], in
particular the #what and #approving-a-release sections. Note the use of
terms "must" (pmc must obey requirements [...]) and "may" (releases may not
be vetoed).
How about re-cutting the release after making sure the tests pass reliably?
Cheers,
Hadrian
[1] http://www.apache.org/dev/release.html
On 01/29/2015 06:17 AM, Gary Tully wrote:
Dan,
the test is mixing stomp and openwire and asserting a jms semantic in
the event of a stomp disconnect without an unsubscribe. The patch
applied broker configuration that makes the test work reliably. Having
a reliable redelivery flag semantic has been evolving for some time.
Stomp users have never had redelivery flag semantics that could be
relied on so they won't care. The full story is in the commit log.
On 28 January 2015 at 17:26, Daniel Kulp <[email protected]> wrote:
I’m more or less -1 until someone can more fully explain the issue with
the failing tests.
The “patch” changes the test, but my question really is whether the
original test code is “correct” and is really exposing some flaw in the
stomp code that the new patch is really just working around. Are users
that are using stomp going to have to make the same changes that were done
in the test in their environments? If so, that’s a bigger issue.
Thanks!
Dan
On Jan 26, 2015, at 4:02 PM, Gary Tully <[email protected]> wrote:
Hi folks,
I've just cut a second release candidate for the long-awaited 5.11.0
release.
This release has more than 120 bug fixes and improvements.
Could you review the artifacts and vote? Especially, it would be great
if
you could test the unix shell script and make sure there's no any
regressions
on the platform you're using.
The list of resolved issues is here:
https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12311210&version=12324951
You can get binary distributions here:
https://repository.apache.org/content/repositories/orgapacheactivemq-1014/org/apache/activemq/apache-activemq/5.11.0/
Source archives are here:
https://repository.apache.org/content/repositories/orgapacheactivemq-1014/org/apache/activemq/activemq-parent/5.11.0/
Maven2 repository is at:
https://repository.apache.org/content/repositories/orgapacheactivemq-1014/
Source tag:
https://git-wip-us.apache.org/repos/asf?p=activemq.git;a=tag;h=f6eb86ee31640427d0f953847f38fcf81a71f9e1
The vote will remain open for 72 hours.
[ ] +1 Release the binary as Apache ActiveMQ 5.11.0
[ ] -1 Veto the release (provide specific comments)
Here's my +1
Regards
Gary
--
Daniel Kulp
[email protected] - http://dankulp.com/blog
Talend Community Coder - http://coders.talend.com
