Mailman 3 May 2002 - Twisted

[Twisted-Python] One time only, I __promise__
by exarkun＠meson.dyndns.org 22 May '02

22 May '02

<radix> exarkun: subscribe AND POST to TPML <exarkun> radix: Why? <radix> exarkun: BECAUSE Yay Twistardeduldoi. ttfn. "There is no reason for any individual to have a computer in their home." -- Ken Olson, President of DEC, World Future Society Convention, 1977 -- 11:38pm up 1 day, 0:17, 0 users, load average: 0.10, 0.19, 0.14

2 2

[Twisted-Python] What we need for next release
by Itamar Shtull-Trauring 21 May '02

21 May '02

1. Switch to new Process API, unix and win32, and write tests for it. 2. Switch to new thread dispatching API. 3. Deal with connector issue. Anything else? (switching to new APIs internally, docs, etc. can be added as we go along.)

3 3

[Twisted-Python] Re: [Twisted-commits] aaaaargh. the makeidentity test was FAILING, and we never noticed, cause Deferred catches exceptions in errback (this fixes failure, not the Deferred issue)
by Andrew Bennetts 20 May '02

20 May '02

On Wed, May 15, 2002 at 11:34:56AM -0500, itamarst CVS wrote: > Modified files: > Twisted/twisted/test/test_cred.py 1.5 1.6 > > Log message: > aaaaargh. the makeidentity test was FAILING, and we never noticed, cause > Deferred catches exceptions in errback (this fixes failure, not the > Deferred issue) Incidentally, I've got a workaround for the errback problem in test_ftp.py. It's not pretty, but it works ok. -Andrew.

2 1

[Twisted-Python] Re: [Twisted-commits] aaaaargh. the makeidentity test was FAILING, and we never noticed, cause Deferred catches exceptions in errback (this fixes failure, not the Deferred issue)
by Christopher Armstrong 16 May '02

16 May '02

So, we need two things for tests: 1) a single reactor that all tests are run in 2) an errback that we can attach to deferreds to make PyUnit show the error (no idea how to do this one) On Wed, 2002-05-15 at 12:34, itamarst CVS wrote: > Modified files: > Twisted/twisted/test/test_cred.py 1.5 1.6 > > Log message: > aaaaargh. the makeidentity test was FAILING, and we never noticed, > cause Deferred catches exceptions in errback (this fixes failure, not > the Deferred issue) -- Chris Armstrong << radix(a)twistedmatrix.com >> http://twistedmatrix.com/users/carmstro.twistd/

3 3

[Twisted-Python] Re: [Twisted-commits] changed reactor module names
by Christopher Armstrong 15 May '02

15 May '02

Isn't it about time we broke all the reactors off into a separate subpackage of twisted.internet? On Tue, 2002-05-14 at 11:50, itamarst CVS wrote: > Modified files: > Twisted/twisted/internet/gtkreactor.py None 1.1 > Twisted/twisted/internet/javareactor.py None 1.1 > Twisted/twisted/internet/pollreactor.py None 1.1 > Twisted/twisted/internet/qtreactor.py None 1.1 > Twisted/twisted/internet/tksupport.py None 1.1 > Twisted/twisted/internet/win32eventreactor.py None 1.1 > Twisted/twisted/internet/wxsupport.py None 1.1 -- Chris Armstrong << radix(a)twistedmatrix.com >> http://twistedmatrix.com/users/carmstro.twistd/

2 1

[Twisted-Python] process.py
by Itamar Shtull-Trauring 13 May '02

13 May '02

I tried refactoring process.Process to support new API, but I gave up - it seems kinda broken, i.e. closes when the process's stdout closes, instead of when the subprocess dies. Anyway, someone else should fix it. I will do the win32 Process. We probably ought to write a few new tests. Once we do that, and rename all reactor modules, I think we're go for a new release.

3 4

[Twisted-Python] Re: [Twisted-commits] Expanded defer.html to explain Deferreds in a little more detail.
by Andrew Bennetts 11 May '02

11 May '02

On Sat, May 11, 2002 at 05:29:22AM -0500, spiv CVS wrote: > Modified files: > Twisted/doc/howto/defer.html 1.3 1.4 > > Log message: > Expanded defer.html to explain Deferreds in a little more detail. > > In particular, I've tried to explain the similarities between try/except and > callback/errback. Criticism welcome :) > > Incidentally, could someone have a good look at the last bit, where I point > out the difference between .addCallbacks and .addCallback + .addErrback? I'd > like to know if > a) I'm describing the current behaviour correctly, and > b) If other people consider this behaviour to be confusing too. Actually, I realised I phrased that question badly. I meant: b) If other people think that the current API docs don't make this behaviour clear enough too? Would anyone mind if I reworked some of the docstrings in defer.py? -Andrew.

1 0

[Twisted-Python] Deferred support in DOMTemplate
by Donovan Preston 11 May '02

11 May '02

Hello everyone A while back glyph suggested I add Deferred support to DOMTemplate. If a specific node handler wanted to defer it's return value, it would create a deferred and return it. He also requested a simple example of using it. I thought it might be interesting to more people than just glyph, so here it is. This will only work with the latest CVS. I might explain a couple of things while I am at it: DOMTemplate is an HTML templating system that uses the DOM (Document Object Model) to give Python code an interface for manipulating an HTML template. Basically, you register your interest in tags containing certain attributes, like a particular class, id, or tag name. Then, DOMTemplate crawls the dom looking for matching tags. When one is found, your callback is called with an instance of xml.dom.minidom.Node, and you are able to create new nodes, move, copy nodes, and delete them. Thus, all of the repeat logic is implemented in Python, and no templating logic is in the HTML at all -- merely attributes on tags that hook them up with Python methods. An rpy is a Resource-generating python file. If you place an rpy in a twisted.web.static tree, the rpy handler will execute the file and look in the resulting namespace for an item named "resource". This gives you a quick and easy way to deploy new applications into an already-running web server. It is also great for debugging, because the file will be reloaded each time you hit it. Hope this is enlightening! Donovan

2 1

[Twisted-Python] a patch
by Jaakko Niemi 11 May '02

11 May '02

Attached is a small patch to make webwidgets shut up or complain when I feed them invalid stuff. What's the policy regarding situations where something gets wrong type of object, for example: traceback or handle it?

1 0

[Twisted-Python] XWT
by Itamar Shtull-Trauring 10 May '02

10 May '02

Hi, I discovered XWT yesterday, and *wow*. This is very similar to something I've wanted to do, and it's very very cool. Now, the thing is, my vision was slightly different - XWT seems to use the web model (XML-RPC and SOAP), where you send a request and get a reply, but there's no way for server to send data to client. A browser is of course limited to this because of HTTP, but in this case you don't need to limit yourself. Why is this a limitation? Consider a chat application. In the XML-RPC model, you would need to continously poll the server for new messages - and if you want low latency, you need to send even more requests per second. This is very inefficent and puts a large load on the server. What I was thinking of therefore was to use a protocol that is two-way, allowing the server to send commands to the client asynchronously. These would basically be GUI events just as you would from, say, Javascript. For example, in a chat application, a new message would result in the server sending a "add new text to message display wdiget" command being sent to the client, without the client having to request it. In addition for event driven client-server applications, using such a protocol would also allow building P2P apps that use XWT. My protocol of choice would be PB, which is a remote object protocol developed for the Twisted server framework (http://www.twistedmatrix.com/products/spread). It has implementation in Python, and a Java implementation I wrote (http://itamarst.org/software/TwistedJava-0.5.tar.gz). PB is very efficient, asynchronous, and two-way (for example, encoding an array containing a short string and two integers will take 150 bytes in XML-RPC, and 20 in PB). It was designed for chat systems, massive multiplayer games and other systems where latency must be kept at a minimum. Given the time I would build such a system - a two-way version of XWT - using Twisted as a server and PB as the protocol. I don't really have the time for this right now :) However, some day I probably will, and it'd be great if I could reuse the XWT code, since most of the code for what I'm envisioning is basically identical. I would therefore be very grateful if your design kept such a future requirement in mind. I'd be even happier if you'd decide to support PB as well, but I'd guess that's less likely to happen ;) Anyway - great work!

3 2