Mailman 3 April 2003 - Twisted

[Twisted-Python] release momentum
by Kevin Turner April 8, 2003

April 8, 2003

Our release schedule, or lack thereof, is getting to me. I'm reluctant to deploy new releases of CVSToys because I'd be asking people to depend on an "alpha" version of Twisted. At least one person has come in to #twisted asking if there was going to be a release soon, or if he had to ask his users to get Twisted from CVS. Our last *alpha* release was was five weeks before PyCon started; there have been over five hundred commits since that time. I know we've all been busy, I know our release manager has had to devote his attention to people who will help him pay his bills -- I'm not concerned with any of that now. All I want to know is what we have to do to get another alpha release out. Tests we're having problems with: * testSignal in test_process.PosixProcessTestCase, fails regularly under buildbot for python2.2, but not 2.1. I cannot reproduce this failure. * testStor in test_ftp, intermittent failures. * ThrottlingTestCase in test_policies has intermittent failures, they may be spurious timing issues. * test_imap is under heavy development, failures changed fresh daily. * test_jelly.CircularReferenceTestCase, Jasper's bug. How many of these do we have to turn green before we can build the next alpha release? Is the jelly bug going to wait on newjelly? Since I'm the one raising a stink, I'll volunteer to manage the next release should Chris remain unavailable. (Now why did saying that start the music of foreshadowing doom in my head?) - Kevin -- The moon is waxing crescent, 6.1% illuminated, 2.3 days old.

6 6

[Twisted-Python] What's twistd doing on Windows?
by Alan McIntyre April 8, 2003

April 8, 2003

Hi all, First of all I wanted to say thanks to all the Twisted guys that gave talks at PyCon 2003. I was going to use Zope for an upcoming project, but the demos you all gave made me think that Twisted might be a better fit for what we're doing. So I had to have a look once I got home. :) So far it's looking good; getting Twisted installed and a web instance running was a snap on both Windows and Debian. I had zero knowledge of Twisted prior to PyCon, but I was able to start from scratch on Saturday and put together a simple web interface to some of our existing Python apps by Sunday afternoon. I did notice something that I'm curious about: I've been running two Twisted web instances, one on a Win2k machine (1.4GHz P4, Python 2.2.2) and one on a Debian machine (266MHz PII, Python 2.1.3) for a few days now, and the Windows instance has accumulated almost an hour of CPU time wherease the Debian instance has only used 2 seconds. What's the Win32 instance doing while it's idle? (It's not a big deal, I'm just curious) Thanks! Alan McIntyre

5 5

[Twisted-Python] FTP Server
by Stefano Maniero April 8, 2003

April 8, 2003

I'm a new developer in twisted. I want test twisted writing an ftp server. Can anyone indicate an example or doc for developing ftp server? In twisted distribution I didn't find it. Thanks

5 6

[Twisted-Python] coding standard: plurals for database table names
by Clark C. Evans April 7, 2003

April 7, 2003

I noticed in the Twisted Coding Standard that... "Database tables shall be named with plural nouns" I'd like to question this standard for several reasons: 1. There is not one single rule for forming plurals. In particular, some plurals have the same spelling as their singluar, 'deer'; furthermore some plurals have different forms, 'people' instead of 'persons'. Sometimes it is 's' or 'es'. These differences are especially hard for those who have English as a a second language. 2. Tables are to rows as classes are to instances. We name Classes singluar, and as such tables should also be singluar. This makes mapping rules easier. 3. It is helpful to name foreign key columns as the same name as the table that they are coming from. And it makes little sense to use plurals for column names, for example, when the person table may have a 1-M relationship with address, it could have a 'address' column name, which is obviously a foreign key to the 'address' table. If you used plurals for the table name it brings up the question... do you use plurals for the column name? I've seen it done. *ick* 4. I often use the double underscore method for link tables which represent M-M relationships. In this case good names are hard to find. For example, a link from a person to their address... how do you name this? I use person__address. If you use plurals then this becomes persons__addresses, dooable but ugly. 5. While there are astechic reasons for this (above) the resons become all the more important when you want to build automatic tools for searching over schemata, etc. In this case, using plurals only adds one extra (unnecessary) step in the automated translation of plural table names to singluar column and or class names. In short, I'd recommend singluar nouns, with M-M link tables as one__two where one and two are tables being linked to. Best, Clark

5 6

[Twisted-Python] enterprise.row
by Justin Ryan April 7, 2003

April 7, 2003

I'm attempting to implement a simple database application (again) using twisted.enterprise.row. Perhaps my failure to do so successfully in the past should be a hint that I should give up, but the concept seems really cool, and I keep getting closer. I'm running into an issue that no doubt has something to do with my lack of understanding of t.e.row. I've been wanting to try it for a while, and decided to design a simple database with the goal of implementing a simple application using t.e.row. In any case, it seems I am angering row. My first attempt at using it was to try and create/insert a new row, and i received the following exception: <- snip -> Traceback (most recent call last): File "Zone.py", line 33, in ? reflector.insertRow(newZone).addCallback(onInsert) File "/usr/lib/python2.2/site-packages/twisted/enterprise/sqlreflector.py", line 308, in insertRow sql = self.insertRowSQL(rowObject) File "/usr/lib/python2.2/site-packages/twisted/enterprise/sqlreflector.py", line 301, in insertRowSQL args.append(self.quote_value(rowObject.findAttribute(column), type)) File "/usr/lib/python2.2/site-packages/twisted/enterprise/row.py", line 87, in findAttribute return getattr(self, attr) AttributeError: ZONE_SOA instance has no attribute 'ZONE_SOAID' <- snip -> "ZONE_SOAID" is the primary key for the "ZONE_SOA" table. I did not set the primary key because it is a postgres serial column - that is, it should automatically increment the pkey each time a new row is inserted. If I use setKeyAttr("ZONE_SOAID", 1), the AttributeError comes up for the next attribute that is not being given a value. I read the row documentation a little more, and it doesn't appear to like what I'm trying to do all too much. I can use python to fill the object with default values, but the database will do that for me, and with less work on the part of my application (plus, the KeyFactory class says it is not thread-safe, I know that my postgres series is). Is there any way for row to allow me to insert a new row with some empty values, and simply ignore them (allowing the database to use its' default)? If not, any idears on where/how this might need to be done? Perhaps this is a good time for me to foray into twisted hacking.. TIA, -Justin

5 12

[Twisted-Python] Twisted systems talking to one another
by Michael McLay April 7, 2003

April 7, 2003

Are there any examples of having two twisted servers exchanging information. The scenario I need to automate is one in which an Original Equipment Manufacturer (OEM) negotiating a contract with an Electronic Manufacturer Systems (EMS) company. The EMS would procure parts and build a product, such as a generic personal computer, a router, or some other type of electronics that are mass produced, but with very short product lifecycles. In this scenario the OEM would use PerspectiveBroker to publish the content of a Request for Quote. The contents of the product data package would not be entirely revealed to potential bidders. The data would be gradually made available based on the state of negotiation in the contracting process. This pattern parallels the gradual disclouse of information about the virtual world in a MUDD game. The trading partners who may wish to bid on the RFQ would copy the initial RFQ announcement to the local server where they could run an analysis program over the basic contract requirements. If the initial requirements look promising they would submit a request for additional information, passing in their qualifications in exchange for the added information. A non-disclosure agreement would be part of this exchange. Then eventually goal would be to respond to the RFQ with a quote and receive a contract to build the OEMs product. Between the initial receipt of the RFQ and the sending of the quote the two servers would engage in a series of moderated transactions for negotiating the terms and conditions for the contract, as well as negotiating the schedule for delivery and the potential to substitute parts in the assembly process. Some of the transactions between the two companies could be totally automated, while others may require human intervention in the decision making process. According to the PerspectiveBroker documents I think this process could be made to work. One problem will be in setting up a universal authentication mechanism. This may just require using LDAP at all trading partnet sites and then federating the LDAP servers into a trading community. I haven't done something like that before, so that might be a bit of a challenge. There are standards for the informationthat need to be used during the transactions (RosettaNet and OAG define that XML representations of documents such as an RFQ; and IPC defines representations for bills of materials of parts to be used and full product data packages), so I won't need to make up that part of the system. The rules governing the negotiations are partially defined in the workflow definitions of the document standards. I think most of the effort will be in building Gtk user interfaces and building the business logic to support the interaction between the trading partners. Each stage in the negotiation process needs to be scripted and a user interface needs to be created. So back to my original question. Are their examples that do something similar, perhaps in a different context. Do TwistedMatrix MUDD worlds talk to each other, do characters and materials shift between worlds, or is each world operating in isolation? I can envision how the operation of the trading partner community could be created as a game, or manufacturing simulation environment. Once the bugs were worked out the simulator could be used to build and manage a real world virtual manufacturing community. Anyone care to estimate how long it might take to build the game version of the system using Twisted? I'm too new to the software to have a clue on what will be involved.

1 0

[Twisted-Python] libivykis: Useful for Twisted?
by Jeff Waugh April 7, 2003

April 7, 2003

"libivykis is a thin wrapper over various OS'es implementation of I/O readiness notification facilities (sich as poll(2), kqueue(2)) and is mainly intended for writing portable high-performance network servers." Useful for Twisted? (spiv was telling me about unreleased epoll goodness, perhaps this will help the Free World?) - Jeff -- GU4DEC: June 16th-18th in Dublin, Ireland http://www.guadec.org/ "Have you ever kissed a girl?" - Bryan Cantrill to Dave Miller

2 1

[Twisted-Python] Copyright issues in the sandbox?
by Bob Ippolito April 7, 2003

April 7, 2003

Two quick legalityish questions: 1) Is it acceptable to commit sets.py (from Python 2.3) into the sandbox 2) I made a pure python (2.2.x+) implementation of itertools, but I stole 99% of the code from the "Equivalent to:" text in the documentation for the C module ( http://www.python.org/dev/doc/devel/lib/itertools-functions.html ), is that legal to put into the sandbox? -bob

1 0

[Twisted-Python] [PATCH] row / reflector / sqlreflector - case insensitive tables/columns
by Justin Ryan April 7, 2003

April 7, 2003

Attached is a patch (cvs diff -u) to twisted.enterprise.row (and related code). Specifically, it affects the following files: twisted/enterprise/row.py twisted/enterprise/reflector.py twisted/enterprise/sqlreflector.py twisted/enterprise/util.py The purpose is to allow for case-sensitive table/column names. For me, there are two parts to this: (1) lots of comparisons in twisted use string.lower (RowObject.__getattr__(), etc..). There, theoretically, shouldn't be any column names that are the same except for their case, but it still seems logical to allow for case-sensitivity. This change depends on the RowObject's case_sensitive_relations attribute (which must be set by child classes). (2) I use postgres, and postgres requires you to enclose mixed-case relation names in double quotes. Thus, twisted needs to quote table names, column names, etc.. This change depends on the sqlreflector's quote_relations attribute, which can be passed as a keyword paramater to sqlreflector.__init__(). I have given this some basic testing, but am mostly posting it for scrutiny whilst I continue to work on it. I think there are some areas that can be improved, but it works and is pretty clean. I think that this would be an appropriate time to state my intent to become the maintainer of the row code, since it has apparently become abandoned. I have several other changes in mind, including the ability to insert new rows without setting their primary keys and such, but I thought that I should start submitting patches for little stuff to make sure that the code I'm writing is sane for twisted. TIA, -Justin

1 0

[Twisted-Python] Re: [Twisted-commits] cleanup after test
by Christopher Armstrong April 7, 2003

April 7, 2003

On Sun, Apr 06, 2003 at 02:30:44PM -0500, warner CVS wrote: > Modified files: > Twisted/twisted/test/test_journal.py 1.8 1.9 > > Log message: > cleanup after test Hmm, what's the point of this? I thought any files you create get stored in the _trial_temp directory (because trial os.chdir()s, or something), which is explicitly left around for post-mortem debugging. Or were these tests creating files explicitly in other locations? -- Twisted | Christopher Armstrong: International Man of Twistery Radix | Release Manager, Twisted Project ---------+ http://twistedmatrix.com/users/radix.twistd/

2 1