programmierung - 4.3.2004 - 6.5.2004

Calerga - SysQuake - Mathematics software for OS 9 and OS X

LispMe Homepage - Lisp for the Palm - also for Palm OS 5

PyLinda - Tuplespaces in Python - this time a fairly complete version

Specification for Fault Code Interoperability - Standard Error Codes in XML-RPC

[I18n-sig] Unicode surrogates: just say no! - Guido van Rossum explains why UCS-4 for internal strings is the best choice in Python

Little Snob - automatic classification of numerical datasets using statistical methods in Python

A, B, C, ... D! The Programming Language - OSNews.com - Description of the D programming language

Alan Kay to receive Turing Award

Great!

I found the original article at Lambda the Ultimate.

iamphet.nm.ru - Scheme stuff (MzVim) - Integrate MzScheme as a scripting language in VIM

dkbza - pydot - Python Interface for the GraphViz DOT language

Zope.org - Readme file for ZopeEditManager 0.9.3 - Integrate External Editor in Zope under OS X

PLT Spy - News

With PLT Spy - the Python implementation for the PLT Scheme runtime - development continues. The current approach is to bring the CPython API to PLT Spy and integrate it into the PLT Scheme runtime, making it possible to use CPython extensions. The goal is to be able to use almost all extensions and also the original C code for standard types in PLT Spy, thereby making all CPython libraries available at once.

That sounds very good, but it raises the question of why one should still use PLT Spy - the linked article provides some answers, but concrete details will only be apparent once things have progressed further.

Here you can find the original article.

Divmod.Org :: Home :: Projects - A bunch of very interesting Python projects

Logilab.org - Aspects - Aspect Oriented Programming with Python

Solution for previous problem

So, I've now simply moved my email validation to an external server, which I query via XML-RPC. It then performs the email validation - and since it runs on a static IP, it's also accepted by AOL.

Anyone who wants to play around with this, the service is accessible via SOAP or XMLRPC. The address for XMLRPC: http://simon.bofh.ms:1111/RPC2 and then call the method mailcheck.validateEmail(adr) there. The address for SOAP: http://simon.bofh.ms:1111/SOAP/mailcheck and then call validateEmail(adr) there. Documentation of the method for this module can be found at http://simon.bofh.ms:1111/API/mailcheck. A WSDL for .NET people and others who need that sort of thing can be found at http://simon.bofh.ms:1111/WSDL/mailcheck. By the way, the web service was created using TooFPy. The corresponding tool is included in the source package - or can be viewed directly in CVS.

Randall D. Beer - FPC-PPC - Floating Point Compiler for OpenMCL and MCL

PyOXIDE - pythonmac.org wiki - Mac OS X IDE for Python - still in development

Various bits of software

Those who enjoy programming with OCaml will find some very interesting and especially practical programs under the link above. I particularly like spamoracle, a very useful filter for statistical spam analysis. I'm now using it on my new server. The original article can be found here.

Translucent Inter-Process Service Migration - Migration of services from one server to another for planned downtime purposes

PyWX: Python for AOLserver - Embed Python in AOL Server as an alternative to TCL

Google Search: - Prototype based programming as metaclass hack for Python

Prothon - Interesting mix of Python and Self

Python Package Index Tutorial - Explanation of how to build PyPI entries in setup.py

Static Type Inference (for Python) with Starkiller

A quite interesting document about a static type checker for Python. It would be nice if something like that would make its way into Python - the possibilities for optimizations by the compiler would be significantly better.

I found the original article at Lambda the Ultimate.

The Guardian hammers RSS

Rogers Cadenhead summarizes what bothers me most about Atom: the silly hype from some proponents who feel compelled to badmouth RSS in an embarrassing way - but ultimately only prove that they simply have no idea what they're talking about. Or if they do have knowledge, they deliberately make false statements, just to promote Atom - a specification that isn't even a draft yet - as the greatest invention since sliced bread.

It's somehow ridiculous how adult people can get so worked up over a stupid file format...

You can find the original article at Workbench.

loaf - Interesting implementation of a Bloom filter

XMLmind XML Editor: XMLmind XML Editor - Interesting XML WYSIWYG editor, for example for DocBook

Tuple Space - Description of Tuple Spaces

Deutsche Zope User Group - Solution for the Compile Problem with DCOracle2 and Python 2.3

First Open-Source License Made for Germany

A nationally adapted open source license is all well and good - but what can it really achieve? Open source is unlikely to be restricted to national borders. And Germany is only a very small part of the world. In my view, it makes much more sense to analyze the existing licenses and work directly with the license holders (such as the FSF for the GPL) to ensure that these general international licenses have full validity everywhere possible.

At most, it would make sense to develop a license addendum that can be attached to the GPL to clarify national peculiarities. Because this way, the exact opposite of the GPL problem could occur: the license is valid in Germany, but not valid outside or only restricted. And that would certainly be just as fatal...

OK, one could do dual-licensing of projects, but then it must be ensured that the licenses don't contradict each other, and that regional restriction is even possible. DFSG, for example, does not allow discrimination by location - and thus a package that is under GPL for international use but under the Bremen License for national use in Germany would potentially no longer be DFSG-compatible.

The original article can be found at heise online news (here).

placenamehere.com projects pnhtoolbar - Web Tools Toolbar for Firebird and Mozilla

PyProtocols - Interface Declarations for Python

myelin: Feed Normalizer

If you have problems reading Atom feeds or broken RSS feeds, you can try Phil Pearson's Feed Normalizer. It can read quite a lot and produces clean RSS 2.0 from it. A very nice tool for a good purpose, especially after that silly solo venture by blogger.com (now at Google).

Here is the original article.

A Busy Developers Guide to WSDL 1.1 - Very compact introduction to the minimal WSDL elements

Affrus 1.0

Wow. That's another one that flew past me. A graphical development environment for Perl. Really graphical, with a graphical debugger, editor, source navigation and everything that goes with it. Pretty impressive. Such a shame that I switched from Perl to Python as my primary programming language some time ago...

Oh, and after I took a quick look at the demo: I think just about everything is right there! The interface starts right up with a graphical shell where you can quickly hack Perl scripts and play around. The graphical debugger is right there attached to it too - if you're into debuggers and stepping through code and looking at variables, you'll be thrilled here.

Actually, there's only one thing I'm missing there: a browser for the Perl modules and the Perl module path and especially the Perl documentation in the modules. Ok, it has some support: if you open the context menu on a module name in a use statement, you can view the documentation in the terminal. And you can load the module into the environment. And jump to directly exported names. But the documentation is only displayed in the terminal or alternatively in the browser. Ok, that's significantly better than nothing, but somehow an integrated documentation browser would really make the whole thing perfect.

In any case, a great environment with a lot of convenience for Perl programmers. Did I mention that I wish something like this existed for Python?

Here's the original article.

Generic SOAP Client - Generic SOAP client with WSDL support for testing

lython - lisp for python - Lisp Frontend for Python

Oracle ...

Q: Why do I get an "ORA-01400: Cannot insert NULL into column name" when inserting a blank string?

A: This is a known Oracle issue. When inserting or updating a value for a varchar2, if you try to insert an empty string (""), Oracle interprets the value as NULL. If there is a NOT NULL restriction on the column in which you are inserting the value, the database throws the ORA-01400 error.

Moppelkotze

zorniges Gesicht

Divmod.Org :: Home :: Projects :: Quotient - Python messaging server based on Twisted

Emmanuel Renieris's Software Page - Python interface to agraph from GraphViz

gnutellavision: intro - Radial layout of graphs

MacNQC - Lego Compiler for OS X and Classic OS

mfGraph Library Homepage - Windows Python module for reading GraphViz Dot files

Parser-SIG - SIG on Parser Generation for Python - Liste von Parser Generatoren für Python

pyparsing -- a class library for text processing in Python - Interessante Klassenbibliothek für Parser in Python

Scalable Vector Graphics (SVG) 1.1 Specification - Vector graphics on the web - Specification

WAVE 3.0 - Web Accessibility Versatile Evaluator - Useful accessibility tester

LWN: The GPL Is a License, not a Contract

Because I'm currently dealing with GPL-FUD in a discussion again, here's the appropriate link to an explanation of why GPL is not a viral license with contractual character, but simply one thing: a license for the use of source code.

Incorrect use of GPL source does not automatically make the new project GPL as well. The reverse path is indeed correct: if you want to use GPL software as an integral component, then your project must also be under GPL. But that effect works in exactly only this direction. Incorrect use of GPL source at worst only leads to one thing: the withdrawal of the license to use that source. So the project simply has to do exactly what it should have done anyway: write the corresponding part itself.

Equally absurd is the claim that you give up your rights to your source code when you contribute it to a GPL project. Of course you retain all rights to your own source code. My experience with arguments of this direction: I'm not contributing to the project because it's under GPL and therefore I can't freely use my own source code anymore are just lazy excuses for what is actually the case: I'm not contributing to the project because I can't freely use your source code for my other projects because my projects shouldn't be GPL. And that's something entirely different - nobody takes away your rights to your own source code (unless you explicitly give them up - in the context of official GNU projects it's customary to transfer rights to the FSF). But you don't automatically gain rights to other source code just because you contributed something yourself. And another thing: people who threaten that their great innovative ideas won't go into a project because it's under GPL usually don't have anything really innovative to deliver anyway. Actually, the number of innovative contributions to projects is minimal anyway - and strangely enough, the people who really deliver innovative parts have the least problems with the license...

Here's the original article.

Python: module inspect - Introspection for Python modules

Mod-pubsub blog - Infrastructure for Publish Subscribe with Websites