Creosote archive for Sun 28 of Aug, 2005

TWISTED Well, the AU Twisted Sprint was fun. I worked on documentation most of the time, and there was some good work on Twisted VFS (it was moved into the main Twisted repository) and Nufox. Unfortunately I didn't get to see much of Sydney, except for watching Luna Park go wizzing past as I was training above it.

Surprise! As blogging becomes mainstream, the ratio of geeks to everybody else is declining, and popularity, influence etc is following the same pattern. Oh well, big deal, we'll still have our niche, over here, where the mainstream isn't looking. Comment

Everyone in the USA seems to think that the USA is way behind everyone else in mobile usage. When I was there last month, though, it seemed that they have a very high level of mobile usage - for voice, at least. Here in NZ, voice calls are pretty expensive, so the main usage of mobile phones is for text messaging.

"PiTiVi allows users to easily edit audio/video projects based on the GStreamer framework..."

Today, the GMail Loader is a graphical, cross-platform, Python based utility that supports two mBox formats (Netscape, Mozilla, Thunderbird, Most Other Clients), MailDir (Qmail, others), MMDF (Mutt), MH (NMH), and Babyl (Emacs RMAIL)

べんりそうだ

OK, this sounds like a truly weird idea, but it might just work: Atom represented as a microformat. This is a continuation of the microformat philosphy: marking up things which already have meaning in HTML. So when you render a post, instead of rendering it like this: <h3>Post title</h3> <p>First post paragraph</p> <p>Second post paragraph</p> You render it like this: <h3 class="title">Post title</h3> <span class="content"><p>First post paragraph</p> <p>Second post paragraph</p></span> This makes it easier for a scraper to figure out that it's a post, and to get the actual data out. Anyone starting to notice that microformats are totally aimed at making life easier for Technorati and other search engines that don't grok RSS and feeds? What I don't like about this format is that it's a total replacement of existing technology - you can already archive your posts in RSS 2.0 format, and the accepted standard for interchange these days is the Movable Type export/import format. That said, a) not everybody archives their posts in these formats, b) only Radio does it for you automatically, and c) adding the microformat tags only requires a small template change in any blogging tool, so it's way easier. One suggestion I'd like to make is that the CSS class names used should be a little less ambiguous.

3D computer graphics utility library for Python.

It provides a threaded web server (you don't have to provide a web server such as Apache yourself), Ypages (Python HTML template language, similar to Java's JSPs) and Snakelets: code-centric page request handlers (similar to Java's Servlets).

Turns out that all of the GLU and GLUT constants were mis-declared. I used the same script to rip them out of the header as I did for GL... but in gl.h the constants are declared in hexidecimal. In the other two they are declared in decimal, and the ...

Excellent! This will only make sense if you live in New Zealand, but now you can make your own National billboard. Comment

I've added the ability to include the solutions when generating sudoku puzzles on my website. The sudoku generator is here. Thanks to Thomas Mills Hinkle for gnome-sudoku.

Hop, skip, and jumping around... I'm taking another look at Oz, a "multiparadigm" language I took a brief look at a few years ago. Along with the book "Concepts, Techniques, and Models of Computer Programming" my first look was brief and did not I did not appreciate everything about them. I was originally put off by the shear number of capabilities in the language and presented in the book.

Just finished the surgery to make OpenGL-ctypes do error-checking for all functions that allow error checking (i.e. those outside of glBegin/glEnd). It's still a bit ugly looking in there (the scar from having two wrapper classes will eventually have...

For a project I&#8217;m working on, I didn&#8217;t want to introduce a dependency on py.test. I also didn&#8217;t want to manually write test suites myself. So, here&#8217;s Testido: a simple unittest extension that easily hooks in to however you&#8217;re testing now and eliminates the need to manually write suites. Through the wonder of setuptools, Testido can be installed trivially with easy_install. The simplest usage looks like this: from setuptools import setup setup(name='foo', version='1.0', packages=['foo'], ) Running &#8220;python setup.py testido&#8221; with that setup file will run all of the tests it can find in the &#8220;foo&#8221; package.

Link Grammar Parser seems like a really cool library: it parses English sentences, and seems to be quite capable. Looking at the sample parsed sentences, it can distinguish good and bad sentences that have completely similar structure (asterisk means it doesn't parse): (more..)

SQLObject 0.7b1 is out. The code is in good shape, I believe, but the packaging and some of the documentation may not be.

Went to put some poems on the texts page and realised that the list of poems was getting a bit long. It also lacked any particular ordering scheme. So I restructured it a bit.

I submitted a patch to Django Ticket #399 (request for a bigint field type). It still needs testing but works at a quick glance on mysql. Here&#8217;s a shot of them in action from the admin interface (the input is just too small and just too big respectively):

generated on Sun 28 of Aug, 2005 at 01:32:20 PM in Melbourne, Australia

this page is generated once a day.