Mailman 3 February 2006 - Twisted

[Twisted-Python] Booksigning Party at PyCon This Year!
by Jeff Rush 05 Feb '06

05 Feb '06

Because many authors of books related to Python will be in attendance at PyCon, we would like to organize a book signing activity. It seems logical to hold it Saturday night at the party at the Nerdbooks.com bookstore. Bring your own books or buy new ones in the store! And if you're an author in attendance who would like to promote your book, drop an email to "jeff at taupro.com" or come forward when we call the book signing to order in the store. So that the bookstore can have copies on-hand, we'd like to collect a list of titles you'd consider having signed. Please edit the following wiki page and add your choices. Authors, you can add your books too. http://us.pycon.org/TX2006/BookSigning -Jeff

1 0

[Twisted-Python] Twisted success story
by christian simms 04 Feb '06

04 Feb '06

As a happy (if that's possible :-) user of Twisted and Nevow, please add the snippet below to the Success Stories page if you'd like. Since I'm delurking now, I might as well 'fess up that I run Twisted on Windows for development (I'm part of the silent mass of Windows users that Glyph was looking for), but I use Linux for production. And since the performance of PostgreSQL sucks on Windows (at least up through PostgreSQL 8.0), I only do heavy duty testing on Linux. I have looked at the skipped Twisted tests on Windows, but have not yet found any that I could fix myself. Though if anyone has suggestions I'd be welcome to devote a little time to helping out. Here's some general feedback: 1. Core of Twisted is really solid. Design is excellent, and almost all problems I encounter have already been taken into account in API. 2. I've gotten burned when trying to keep synchronous-style code synchronous, like by using the naughty deferredResult. I did it (just in my admin tool) even though you guys say not too. The only way to learn not to do it is to touch the flame and get burned. 3. Nevow is still moving a lot, but you guys know that. Parts of it are as good as Twisted. Even though I initially loved stan, I got wacked by its performance and had to move to more string based rendering, since my app is really dynamic I can't use standard type templates at all. Keep up the good work! Cheers, Christian --Begin-- Botonomy LLC is a small software firm in the Philadelphia area. We help small teams solve large problems. Our first application, ProjectPipe (http://www.projectpipe.com), is a hosted project management solution that provides midsize teams with everything that they need to manage IT projects. We have built our hosted application infrastructure atop Twisted, Nevow, PostgreSQL, and Jabber/XMPP. Twisted's multi-protocol support has been one of the key enablers of the unique technical architecture that we employ in our product offerings. The core Twisted development team consistently demonstrates a strong commitment to delivering robust, high quality software. Twisted's documentation is accurate and concise, and the expertise and helpfulness of the larger developer community is second-to-none. When evaluating the technologies on which you are staking your business, you need to evaluate both the code and the community that surrounds it. Twisted passes with flying colors in both regards. --End--

2 1

[Twisted-Python] What happened with API reference?
by Alec Matusis 04 Feb '06

04 Feb '06

When I try going to http://twistedmatrix.com/documents/current/api/ I get a weird site, not the API reference I used to see. Am I missing something?

2 1

[Twisted-Python] Running callbacks in a non-main thread
by Gabe Rudy 04 Feb '06

04 Feb '06

Hey, I realize that the point of deferreds is to have process/network intensive tasks work on in their own thread and send the results to callback function in the main thread, but is there away to have a deferred process it's callback in a non-main thread? I'm sure someone will want to know why, and maybe solving my meta-problem would show me the errors in my ways in wanting to warp the purpose of the mighty deferred. Async is great, but I'm slowly replacing some code that uses xmlrpc (blocking) with prospect broker objects (returns deferred, hence non-blocking). What I'm trying to hack is a method to make a non-blocking call in the main thread blocking until the deferred returns a result, hence emulating the blocking nature of xmlrpc. This dependency is something I would like to phase out, but certain things need it, so work with me eh? My hack is to simply use a threading.Event to wait for the callback to be called in a non-waiting thread, save the value and do the let the main thread return the result. Of course this doesn't work when callbacks run in the main thread because the main thread is blocked. #Bits of hack code: class Block def waitFor(self, function, *args, **kwargs): if not hasattr(self,'async_done'): self.async_done = threading.Event() self.result = None #ceate a thread to wait for callback. threads.deferToThread(function, *args, **kwargs).addCallback(self._cbWait).addErrback(self._ebWait) #block on results of thread to be set self.async_done.wait() self.async_done.clear() return self.result #This is never called, because main thread is blocked def _cbWait(self, deferred): #expect original function to return a deferred deferred.addBoth(self._cbValue) def _ebWait(self, reason): print "ERROR in getting remote function: ",reason def _cbValue(self, v): self.result = v self.async_done.set() Any ideas O masters of the twisted? Gabe Rudy

5 5

[Twisted-Python] Re: Making DeferredList look like Deferred
by Lenny G Arbage 04 Feb '06

04 Feb '06

Jean-Paul Calderone exarkun at divmod.com Fri Feb 3 14:08:54 MST 2006 wrote: > > DeferredList has *either* callback *or* errback > called on it, just like any Deferred. If you just > give it a list of Deferreds, it will always have its > callback method invoked and never have its errback > method invoked, regardless of whether any of the > Deferreds it was constructed with fail. However, if > you pass a true value for fireOnOneErrback when > constructing your DeferredList, any Deferred which > errbacks will cause the DeferredList to errback; in > this case it will never callback. > > Does that help? That definitely helps, but I'm wondering if this behavior was implemented between 1.3 and 2.0? I can provide an example where both the DeferredList's errback and callback are called on 1.3 when any one of the DeferredList's Deferreds err -- without setting fireOnErrback. I'll grab 2.1 tonight and see if I can reproduce it there. Thanks, Lenny __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

[Twisted-Python] Re: [Twisted-commits] r15819 - Maybe make this test more robust. Windows sucks.
by Jean-Paul Calderone 04 Feb '06

04 Feb '06

On Fri, 03 Feb 2006 00:30:41 -0700, Jonathan Lange <jml(a)wolfwood.twistedmatrix.com> wrote: >Author: jml >Date: Fri Feb 3 00:30:40 2006 >New Revision: 15819 > >Modified: > trunk/twisted/trial/test/test_reporter.py >Log: >Maybe make this test more robust. Windows sucks. > > > >Modified: trunk/twisted/trial/test/test_reporter.py >============================================================================== >--- trunk/twisted/trial/test/test_reporter.py (original) >+++ trunk/twisted/trial/test/test_reporter.py Fri Feb 3 00:30:40 2006 >@@ -42,6 +42,7 @@ > def test_timing(self): > the_reporter = reporter.Reporter() > the_reporter._somethingStarted() >+ time.sleep(0.01) > the_reporter._somethingStarted() > time.sleep(0.01) > time1 = the_reporter._somethingStopped() I think robustness is inversely correlated with number of time-related calls. What are these sleeps for? Maybe the reporter's timing function should be parameterized so this can be deterministic, instead of relying on a clock that is guaranteed to say something nasty from time to time? Jean-Paul

2 1

[Twisted-Python] Making DeferredList look like Deferred
by Lenny G Arbage 04 Feb '06

04 Feb '06

I have a need to return a DeferredList or a Deferred from a function, but I don't want the caller to be aware of this distinction -- the caller should just assume that they got a regular Deferred. The way I'm currently accomplishing this is to 1) ensure that each Deferred in the List has its own callback/errback pair hooked up, and 2) add a callback to the Deferred list from within the called function which either returns the [possibly sanitized] result if all the Deferreds called their callbacks, or that raises a failure.DefaultException if any of the Deferreds called their errbacks. This seems to give me the desired result in all cases. Is this the right way to go about this? Is there a cleaner way? (The reason I can't just return the DeferredList, as I understand, is that its callback is *always* invoked, even if we leave each of the Deferred's errbacks unhooked and let them trickle up into DeferredList's addErrback callable. In the case of a caller who expects the returned deferred to be a regular deferred, it would not be expected to have *both* its callback and errback invoked). thanks, Lenny __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

[Twisted-Python] callback and errback
by Michele Petrazzo 03 Feb '06

03 Feb '06

Hi list, I have a "stupid" problem with the adbapi (deferred): d = self._db.runQuery("SELECT * FROM not_exists_table") d.addCallback(self._back_getTables, callBack) d.addErrback(self._errBack, errBack) File "C:\Python23\lib\site-packages\twisted\internet\defer.py", line 294, in _startRunCallbacks self._runCallbacks() --- <exception caught here> --- File "C:\Python23\lib\site-packages\twisted\internet\defer.py", line 307, in _runCallbacks self.result = callback(self.result, *args, **kw) exceptions.TypeError: passthru() takes exactly 1 argument (2 given) Why with addCallback I can pass my values and with addErrback no? And if I use: d.addCallbacks(self._back_getTables, self._errBack, callbackArgs=callBack, errbackArgs=errBack) File "C:\Python23\lib\site-packages\twisted\internet\defer.py", line 294, in _startRunCallbacks self._runCallbacks() --- <exception caught here> --- File "C:\Python23\lib\site-packages\twisted\internet\defer.py", line 307, in _runCallbacks self.result = callback(self.result, *args, **kw) exceptions.TypeError: passthru() argument after * must be a sequence What this mean? OS: win2k/debian 3.1 py: 2.3.5 twisted: 2.1.0 The error happen on both OS Thanks, Michele

1 1

[Twisted-Python] small error in the documentation/howto
by Andy Leszczynski 02 Feb '06

02 Feb '06

Hi, I am not sure where to post it, but there is a small error standin out in the: http://twistedmatrix.com/projects/core/documentation/howto/defer.html Basically I believe that in the first frame following change is needed: -def getDummyData(self, x): +def getDummyData(x): A.

2 2

[Twisted-Python] (no subject)
by Wolfgang Eichler 02 Feb '06

02 Feb '06

I don't get the docu for class TcpServer when I do: pydoc.py -w base The class Echo is well documented. For TcpServer I only get: TcpServer = <class base.TcpServer> base.py: -------------- import twisted.protocols.basic from twisted.internet.protocol import DatagramProtocol, Protocol, Factory class TcpServer(twisted.protocols.basic.LineReceiver): """ central class for incoming tcp messages """ def connectionMade(self): """ new connection """ class Echo(DatagramProtocol): """ Handles incoming datagrams (udp, traps) """ def datagramReceived(self, data, (host, port)): """ contains some standard prints tbd """ --------------- Thanks for any hint ...

1 0