Mailman 3 - Twisted - python.org

[Twisted-Python] twisted.net refactoring
by Moshe Zadka 17 Apr '01

17 Apr '01

Hi! I've started to work on refactoring twisted.net for minimal dependancies. Currently, in the vein of "research project", what interests me is how layering of protocols in an event-oriented framework can be accomplished. Please have a look at http://moshez.org/protocols.tgz I've written tests for all the protocol handlers I wrote, and I'd be happy to work with anyone who wants to write tests for the irc client and telnet server. Basically, the idea is to turn low-level events (basically just handleData) into high level events (like handleRequest, for HTTP, at the other end of the spectrum). The idea is to reflect the layering of protocols (and only the layering of protocols -- read on) in the classes. So, inheritance happens when the spec references another spec -- e.g., XML-RPC is defined to be over HTTP, so it would be correct to have class XMLRPCHandler(HTTPHandler): def handleRequest(self, command, selector, version, request): ...parse request into Python tuple... result = self.handleMethod(params) ....return HTTP response... But, on the other hand, HTTP does not say anything about TCP/IP, so HTTP should *not* be layered above any transport. This allows us to use SSL, SOCKS or files just as well. One big advantage I'm interested in is testability -- see the examples already there to get my point -- you can bombard your handler hard-coded strings and watch for the answers. -- "I'll be ex-DPL soon anyway so I'm |LUKE: Is Perl better than Python? looking for someplace else to grab power."|YODA: No...no... no. Quicker, -- Wichert Akkerman (on debian-private)| easier, more seductive. For public key, finger moshez(a)debian.org |http://www.{python,debian,gnu}.org

2 2

[Twisted-Python] weird problem with subclassing library.door.Door
by Chris Armstrong 14 Apr '01

14 Apr '01

Ok, I made a subclass of Door called SecretDoor. The point of this door is that it's opened with a remote button and is not visible unless it is open. I tried implementing this by overriding _close() in my class. class SecretDoor(door.Door): def verb_open(self, sentence): sentence.subject.hears("This door is opened elsewhere.") def _close(self, actor): self._undescribe() Unfortunately, this did not work. The _undescribe call was apparently being cancelled out by something that was resetting the description in the current room. I looked for calls to _redescribe(), but couldn't find any that were being called after my _undescribe(). Eventually I just decided to hack it out and added: def _redescribe(self): if not self.isOpen: self._undescribe() else: door.Door._redescribe(self) This is very bad, but I could not figure out any other way to do it. I still have it this way, and I just wanted to notify glyph and anyone else who's interested. Is there a better way to do this? Where is the description being added back into the room? Ahh well, thanks a bunch. -- Chris Armstrong carmstro(a)twistedmatrix.com http://twistedmatrix.com/~carmstro carmstro(a)dynup.net

1 0

[Twisted-Python] config interface
by Chris Armstrong 11 Apr '01

11 Apr '01

I've been working a whole lot on the config interface, and I don't like how things are working right now. configurable objects are organized in a tree. Every object defines these methods: #ContainableTypes should return a list of classes that #it can hold. def configContainableTypes(self): pass #Type should return the type of the current object def configType(self): pass #GetContents should return a dict of children of this object. {"name": obj} def configGetChildren(self): pass #AddItem should add a child to this object def configAddItem(self, id, item): pass #RemoveItem should remove a child from this object. def configRemoveItem(self, id): pass I'd like to change this API somewhat to allow different paramaters to add and remove item methods. Not all objects need ids to add an object, and it's not natural for some objects to represent their contents as a dict (selectors, for instance). I was thinking about a solution similar to the improper-state trick that glyph described to me, only it would just look at the arguments that a method needs rather than a constructor of a class. So, maybe there could be standardized names for paramaters of AddItem or RemoveItem so that a front-end would know what exactly to get from the user to add or remove an item. GetChildren could also just return either a list or a dict, and the front-end would handle it appropriately. This is not a developed idea at all, and I just wanna run it by anyone who cares. Standardizing names of paramaters doesn't seem very clean to me, but I can't think of much else.. -- Chris Armstrong http://twistedmatrix.com/~carmstro carmstro(a)dynup.net There is a 90% chance that this message was written when the author's been awake longer than he should have. Please disregard any senseless drivel.

2 4

[Twisted-Python] And Now I Eat humble.py
by Glyph Lefkowitz 11 Apr '01

11 Apr '01

Well, the "Super-Duper It-Can't-Break No-Really-It's-Not-Like-The-Java-Version-I-Swear" version of containment in Twisted Reality broke rather nastily after dash sent out his update email regarding the demo. I *think* I've got it mostly fixed now... at least, a nastily contrived, pathological unit test passes, and it appears to do the correct thing interactively. It doesn't try to keep track of nearly so much stuff now, so even if it does break, it's likely that the breakage will be in event updates and not the actual structure of the game. Curiously enough, all this stuff broke when we introduced a certain mechanical cockroach into the demo world... In other news, I've ported the old Java Faucet to Jython (only works on 2.1a1, since previous versions of jython don't implement getattr properly!!), which means that the faucet is runnable with a nice gui on MacOS X now. The other important note is that since Gloop now works on Jython, it is possible to hook arbitrary Java programs and libraries into a Gloop distributed network (this turned out to be very simple, actually). The possibilities of this are left as an exercise for the reader. Currently, I'm adding features to the demo -- testing guest login -- and taking some screenshots of various GUI things. The way things are going at this point, I think the next release (after the demo is done) should be 0.9.0; post that, (after carmstro's config interface is done, and hopefully Inheritance) we should move on to 0.99.0 and concentrate mainly on writing tests... once we've tested every module, 1.0. This concludes this morning's State of the Matrix Address :) -- ______ __ __ _____ _ _ | ____ | \_/ |_____] |_____| |_____| |_____ | | | | @ t w i s t e d m a t r i x . c o m http://twistedmatrix.com/users/glyph

2 1

[Twisted-Python] The King of Spain never rushes
by Allen Short 07 Apr '01

07 Apr '01

AUSTIN (AP) -- Work proceeds apace on the Twisted Reality Demo Center, the long-awaited spectacular showcase from Twisted Matrix Labs. "Eat your heart out Verant, Bioware, and Blizzard", chairman and Evil Overlord Glyph Lefkowitz was overheard saying, "we have more internet than you can possibly imagine. Heck, we've even gotten an endorsement from John Romero." When asked to comment, Mr Romero squeakily responded "That tickles!". Chief Reality Engineer Allen Short refused to discuss specific technical advances over previous releases, saying only "You know it's gonna rock. Where or when are minor details." All seriousness aside, I've begun to port the Java code for the TR demo to the current system. Right now, we have Mr Romero and the water fountain; their code is in trdemo/trdemo.py, along with trdemo/build_map, the mapfile. I've added a Box class to twisted.library; problem is, it doesn't work quite right. (Try putting something in a box to see what i mean =) PS: glyph, have you added the verb-chaining stuff yet? i've a case or two that wants it. -- Allen Short, CS Geek shortwa(a)auburn.edu favorite toys: Lisp/Smalltalk/Python/Jabber/XEmacs/Linux <demoncrat> I think nowadays sisyphus would have to reinstall windows all the time

1 0

[Twisted-Python] upon further consideration...
by Glyph Lefkowitz 02 Apr '01

02 Apr '01

Nick, I think you're right. I'm going to scrap the verb-oriented dispatch system for something MUCH more like Inform's style. I'm still working out the design, but when I've got something substantial, I'll post again. ______ __ __ _____ _ _ | ____ | \_/ |_____] |_____| |_____| |_____ | | | | @ t w i s t e d m a t r i x . c o m http://twistedmatrix.com/users/glyph

2 1

[Twisted-Python] ANNOUNCE: TwistedPython 0.8.0a released
by Glyph Lefkowitz 26 Mar '01

26 Mar '01

0.8.0a is a feature enhancement release focusing on twisted reality. Numerous bugfixes and more documentation are sprinkled throughout the codebase, however, so download and enjoy! ______ __ __ _____ _ _ | ____ | \_/ |_____] |_____| |_____| |_____ | | | | @ t w i s t e d m a t r i x . c o m http://www.twistedmatrix.com/~glyph/

1 0

[Twisted-Python] Inform
by Nick Moffitt 24 Mar '01

24 Mar '01

Is there any push to port the inform library over to twisted-python in its entirety? It strikes me that the current "ruins" demo is lacking in many features simply because the library is so spare (for instance, "get all" is not implemented). I'm interested in this largely because Inform is non-free software, and I'd like to get back into writing interactive fiction. Inform itself was never more than just a simple bytecode compiler and object system, while the library was its true benefit. -- You are not entitled to your opinions. 01234567 <- The amazing indent-o-meter! ^ Matt McIrvin: the Nikola Tesla of tab damage.

2 6

[Twisted-Python] WWW.TWISTEDMATRIX.COM
by Christine Hall 17 Mar '01

17 Mar '01

1 0

[Twisted-Python] some demo scripts
by Norbert Klamann 14 Mar '01

14 Mar '01

Hello all, i started playing with twisted python 0.7.2 and it looks very interesting. I tried to go along the 'philosophical tutorial to Twisted Python ' and created 4 little python-scripts on the way which initialize and use the minimalist telnet- and web-server. They run under windows nt and python 2 but should be portable. Under Windows NT a start of the servers gives an attribute error because os.getgid() is nit implemented. I suppressed this message under NT : line 975 sqq on \twisted\net.py could read : except: if os.name == 'nt': pass else : traceback.print_exc(file=sys.stdout) print 'setuid/gid failed' Enjoy Norbert Klamann Norbert Klamann ++ ++ Klamann Software & Beratung GmbH ++ Franz-Marc-Str.174 ++ D-50374 Erftstadt ++ Tel.: 02235 / 985272 ++ Fax: 02235 / 985273 ++ Mobil : 0172 / 2797723 ++ PGP Fingerprint : ++ 05CC 4362 E765 51F1 38D9 C43F D907 B4D8 7873 3454 from twisted import threadable threadable.init(0) from cPickle import load selector = load(open('twistd_web.spl')) from twisted import web myTest = web.Test() selector.servers[0].root.putChild('test',myTest) selector.run() from cPickle import dump dump(selector) from twisted import threadable threadable.init(0) from cPickle import load selector = load(open('twistd.spl')) selector.run() from twisted import threadable threadable.init(0) from twisted import net selector = net.Selector() from twisted import web server = web.Server(8181, selector) documentRoot = web.File('.') server.setRoot(documentRoot) server.startUp() from cPickle import dump dump(selector, open('twistd_web.spl','wb')) from twisted import threadable threadable.init(0) from twisted import net selector = net.Selector() from twisted import telnet server = telnet.Server(4040, selector) from cPickle import dump dump(selector, open('twistd.spl','wb'))

2 1