Mailman 3 July 2011 - Twisted

[Twisted-Python] How do twisted and multiprocessing.Process create zombies?
by Sean Ochoa July 5, 2011

July 5, 2011

Using twisted loopingcall, multiprocessing.Process, and multiprocessing.Queue; is it possible to create a zombie process. And, if so, then how? -- bitcycle

2 2

Re: [Twisted-Python] [Bug 775213] pymsnt 0.11.3-5 fails to start due to broken python-twisted-words 10.2.0-1 (AttributeError: 'module' object has no attribute '_parse')
by Glyph Lefkowitz July 5, 2011

July 5, 2011

On Jul 4, 2011, at 9:48 PM, Diane Trout wrote: >> Almost - I should have mentioned <http://twistedmatrix.com/trac/wiki/TwistedDevelopment> and <http://twistedmatrix.com/trac/wiki/ReviewProcess> first. >> >> When you put that branch on Launchpad, you'll need to put #4799 into review as well, as per that process; just link to the branch. Also, make it clear that you've volunteered for this work somewhere public, ideally in a ticket comment. >> >> Thanks again! Please don't hesitate to get in touch if you need any further guidance. > > Ok I have linked to my branch in http://twistedmatrix.com/trac/ticket/4799 > > Does that look right? is the next step to put "review" into 4799's keywords field? The next step is actually to make a backported patch specifically for <http://tm.tl/4799>, and get that merged into a release branch. The idea with the release ticket is that it's for the release notes and final QA; all the actual code should be complete in individual feature tickets first. I'm sorry - now I'm noticing that the workflow on the linked tickets (for example <http://tm.tl/4786>) is actually wrong for a maintenance release, as there's no cloned bug to indicate the backport. If I recall correctly, /branches/releases/release-10.2.0-4651-2 is the branch to merge things into for a 10.2.x maintenance release. At this point it would probably be good to get on the mailing list <http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python> and discuss these issues more publicly; someone else with more experience with the release toolchain can provide you more specific advice, and you'll get faster responses than from me personally. I'm Cc:ing this message to that list so that there will be some context when the discussion starts. I was going to recommend the IRC channel too, but I am told you've already been there; I hope I'll run into you there :). > Also I don't know how to get the code changes out of launchpad and into your svn repository. A committer will do that, based on your branch. You may have to use our all-branches mirror to continue work, as Launchpad only mirrors trunk. Or, you may become an committer; we normally ask that people fix a few smaller bugs first, but given that you're volunteering to manage a release we may expedite that process :). > Diane Once again I _really_ appreciate your time here. As you can probably tell, we rarely have the contributor bandwidth to do maintenance releases and so we're somewhat rusty at it. We like to provide a smooth upgrade process for our users though, so when we manage to get this pushed through it will be a big benefit to our XMPP users. Thanks, -glyph

1 0

Re: [Twisted-Python] Moving Twisted off Trac and SVN to somewhere nicer
by Michael Thompson July 3, 2011

July 3, 2011

On Jul 2, 2011 10:28 p.m., "Glyph Lefkowitz" <glyph(a)twistedmatrix.com> wrote: > > > On Jul 2, 2011, at 1:07 PM, Michael Thompson wrote: > >> On 1 July 2011 18:38, Glyph Lefkowitz <glyph(a)twistedmatrix.com> wrote: > > >> I think github means less effort for the reviewer because they can >> review commits, rather than a large diff. They can review changes >> following a review. > > > I already review diffs this way, with a local bzr-svn branch using 'bzr merge; bzr qlog'. Better yet I just sometimes do this on a plane :). > Is that reviewing Svn branches what about text patches? If I address three review comments sending in three patch files seems like a hassle for me and the reviewer and could lead to more delays in the review process.

1 0

[Twisted-Python] Weekly Bug Summary
by exarkun＠twistedmatrix.com July 3, 2011

July 3, 2011

1 0

[Twisted-Python] Conch Testing server and client
by Anshul Singhle July 1, 2011

July 1, 2011

Hi, I have to write tests for a custom conch server. My approach is to make a test client and then connect it to my server in my testcase. I have been able to create a client that can interface with my server, but i'm not able to do it from within the test case. What I'm trying is : class TestAccountStatus(unittest.TestCase): def setUp(self): ssh_server = Server() self.server = reactor.listenTCP(ssh_server.port, ssh_server.application(), interface=ssh_server.interface) def tearDown(self): server, self.server = self.server,None return server.stopListening() def test_1(self): d = protocol.ClientCreator(reactor, SimpleTransport).connectTCP('localhost', self.server.getHost().port) return d The transport SimpleTransport is my custom transport.SSHClientTransport instance. The class Server() is a wrapper around an SSHFactory insatance. The problem I'm facing is that after this line > d = protocol.ClientCreator(reactor, SimpleTransport).connectTCP('localhost', self.server.getHost().port) I have to write reactor.run(), otherwise the client is not starting. If I write reactor.run() the test doesn't stop (understandable since the reactor isn't being stopped). Basically I want to know how I can start the client without running reactor.run() Thanks in Advance!! Anshul

4 8