Mailman 3 July 2009 - Twisted

[Twisted-Python] (Mis)indented Posts
by Gerrat Rickert 27 Jul '09

27 Jul '09

...ok, I've created a new post twice on this list twice, and both times my post has been indented and placed under a completely different/irrelevant post Here: http://twistedmatrix.com/pipermail/twisted-python/2009-June/019859.html And Here: http://twistedmatrix.com/pipermail/twisted-python/2009-July/019902.html ...if this post hijacks a thread as well, then it will be 3 times. ...this seems to be something specific to the twisted python mailing list: http://www.mail-archive.com/mailman-users%40python.org/msg53996.html Gerrat

3 8

[Twisted-Python] Weekly Bug Summary
by exarkun＠twistedmatrix.com 26 Jul '09

26 Jul '09

1 0

[Twisted-Python] Volunteering for Release Manager
by Jonathan Lange 25 Jul '09

25 Jul '09

Hello all, With some trepidation and a complete willingness to fail, I'm offering my services as Twisted release manager, starting Tuesday next week. I have only two questions: - Am I accepted? - Where do I begin? jml

3 4

[Twisted-Python] laxdb - an async dbapi wrapper
by Vishal Shetye 23 Jul '09

23 Jul '09

> Hi, > > In response to the announcement of the Twisted Asynchronous Database Api > (tada), <http://www.darkarts.co.za/projects/tada/, I thought I should > mention something I wrote. > > [Unfortunately, tada does not use threads correctly so it's likely to > break. The author knows this and has added a warning to the module for now.] > > A while back, I started messing around with making a standard, blocking > db-api module appear non-blocking but with a deferred API. I called it > laxdb and it's in my sandbox. > > svn://svn.twistedmatrix.com/svn/Twisted/sandbox/mg/laxdb.py > http://svn.twistedmatrix.com/cvs/sandbox/mg/laxdb.py?view=markup > > laxdb is quite literally an async db-api interface and not a > t.enterprise.adbapi style interface. In particular, it provides exactly > the same API (but deferred) and it doesn't take over transaction > management. It also doesn't wrongly call a cursor, "Transaction" ;-). > > I recently went back to laxdb and finished it off, also adding a > connection pool implementation. > > Now, I haven't used laxdb in a real project and there are no unit tests > but I think it's reasonably complete. It should work with any dbapi > module that has threadsafety level 1 or higher, just like adbapi. > > If there's interest in this way of accessing a database then I would be > happy to write tests etc, move it into Twisted and maintain it. > > Cheers, Matt I guess there have not been any additions to this module in recent time. I just tried to use it, sample code, actually a failure test, is as below which does not actually do any query. Reactor is stopped without commit or rollback. This raises an exception in removeTrigger_BASE. exceptions.ValueError: list.remove(x): x not in list I could fix it, in class WorkerThread, function _start has line self._shutdownTriggerId = reactor.addSystemEventTrigger('after', 'shutdown', self.shutdown) self.shutdown is called by Connection.close() or reactor.stop(). In case of second, reactor.stop, registered function, self.shutdown in this case, should not do reactor.removeSystemEventTrigger(self._shutdownTriggerId) However, this raises some doubts about overall code as there are no test cases to prove the correctness. Has anyone used this before? Can somebody provide pointers on what scenarios/conditions should be tested? laxdb is good match for what I want, interleaved calls and async interface to db. Is there any alternative available? [code] from laxdb import ConnectionPool from twisted.internet import reactor pool = ConnectionPool("sqlite3", database='laxdbTest', timeout=5.0, check_same_thread=False) def connect(pool): d = pool.connect() d.addCallback(connected) def connected(conn): curs = conn.cursor() reactor.stop() reactor.callWhenRunning(connect, pool) reactor.run() [/code] Regards, vishal

1 0

[Twisted-Python] Commit access
by Jonathan Jacobs 23 Jul '09

23 Jul '09

Hello Twisted, I've been contributing to Twisted (jonathanj on the Trac instance, k4y on IRC) for around a year now. I'd be grateful for commit access to the Twisted repo and, as per glyph's suggestion, I'd like to formally request it. I'm not sure if I'm supposed to bow or not, so I'll just exit left. -- Jonathan

2 1

[Twisted-Python] An alternative Twisted GPS protocol implementation
by Laurens Van Houtven 22 Jul '09

22 Jul '09

Hi everyone :-) I'm writing a GPS-based app and I was kind of unhappy with the existing GPS implementation. It probably works, but it has two problems that you might care about if you're using it: 1. it's interface isn't very friendly (twisted.protocols.gps.nmea.NMEAReceiver.decode_fix has _fourteen_ positional arguments, which are essentially raw sentence data -- this reminds me of MFC too much for me to stay in my comfort zone) 2. it doesn't seem to have any unit tests, so behavior is up in the air a little Also, if you want to to extend it (admittedly not very likely with a rusty protocol like NMEA), well, the current implementation is kind of terse. Anyway, I've been working on a friendlier interface for a few days now, and if people like it, it might be a candidate for inclusion in Twisted itself. This would mean deprecating twisted.protocols.gps in favor of the shiny new twisted.protocols.positioning. I was originally going to make it a Twisted community project, but glyph expressed interest in putting it in Twisted itself instead. It does nice conversions like parse the timestamps into Python time/date/datetime objects and automatically adding sign information for coordinates, stuff like that. Also, it provides a much nicer interface to the information, so you can use code like: point.position.latitude point.position.longitude.inDegreesMinutesSeconds point.altitude point.altitude.aboveMeanSeaLevel.inFeet point.heading.trueHeading Et cetera. Use of this interface is optional, there's also straight dict access (which is roughly like the current API except with a dictionary instead of positional arguments). Right now it only supports NMEA. The old system supported Rockwell as well. What I'd like to hear from you (you mostly being potential users of this piece of code): do you care about Rockwell (aka Zodiac) support? If you're just using twisted.protocols.gps now, I'm interested in hearing from you as well of course :-) SiRF support for now is probably out of the question. It's a whole different ball game. Parsing it is *hard* (it's much more complicated than just a binary representation of the same data). Replies welcome, alternatively you can usually catch me at this email address or as lvh on freenode (in the usual suspects: #python, #twisted...) Thanks for-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Use GnuPG with Firefox : http://getfiregpg.org (Version: 0.7.6) iEYEARECAAYFAkpjdcMACgkQT5v5zGkvKT5EpwCeLhrOe1tHXuJQKc78jAQ+ALQA s5IAn3vn5kVEZ++eUGuRb3zsxynQPCR4 =7Cyx -----END PGP SIGNATURE----- reading, Laurens

2 2

[Twisted-Python] Incorrect "Another twistd server is running..."
by Peter Cai 22 Jul '09

22 Jul '09

When my server reboot after an unexpected power cut, supervisord tried to start a twisted app after an unexpected power cut, but it fails because "Another twistd server is running, PID 4607" 4607 is the PID of the twisted app before power cut. After I stopped and restart supervisord, everything is OK. What might cause twistd to think that an non-exist process is running? -- look to the things around you,the immediate world around you, if you are alive,it will mean something to you ——Paul Strand

4 6

Re: [Twisted-Python] Twisted words: Oscar
by Paul Swartz 20 Jul '09

20 Jul '09

On Tue, Jul 14, 2009 at 2:26 AM, Nida <crazyturkey(a)gmail.com> wrote: > Hi there, > > I noticed that you are on the twisted words mailing list and it says that > you are the maintainer too, so I was hoping I could ask you a few questions: > I would like to use the twisted words oscar protocols. > I am trying to use the oscardemo.py example as a guide and I would like to > use sendMessage without call back and I don't want to use the reactor > (because I don't want it to loop). > I would like the bot to log in, but I want some user interaction to type in > a message to send to others. My problem is I want to use sendMessage as well > as getting a buddyList. > Is there a better way then using ClientCreator? Or a simpler way to log in? > (I'm cc-ing the twisted list, since that's the place where questions should go) ClientCreator is just an easier way to connection client protocols, without having to make your own Factory. You have to use the reactor; that's how the TCP data gets to your client. I'm not sure what "I don't want it to loop" means, but you can always stop the reactor after you're done with whatever networking you're doing. You don't need to have a callback for sendMessage. Just don't pass the wantAck kwarg to the method, and you won't get a Deferred back. -p -- Paul Swartz paulswartz at gmail dot com http://paulswartz.net/ AIM: z3penguin

1 0

[Twisted-Python] Timers in Twisted
by David Guerin 20 Jul '09

20 Jul '09

Hey all, I'm pretty new to Twisted Framework but things are going along nicely. At the moment I am trying to get a timer to fire off say every 5 minutes. What's the best way to do this with twisted? I found an old link to a how to with timers in twisted on the twistedmatrix.com site but there wasn't a web page there when I clicked Thanks Regards David

2 2

[Twisted-Python] LoopingCall at a non-idling reactor
by Ilya Etingof 20 Jul '09

20 Jul '09

Dear Twisters, I can't get Twisted reactor calling my function in a more or less periodic fashion with the LoopingCall mechanics. With select()-based reactor, the LoopingCall object seems to call back my timer function only on select() timeout. If I/O pace does not allow select() to time out, LoopingCall never calls my function. How do I implement a "timer" function what would work regardless of the rate of I/O events? thanks, ilya

4 14