The case for RSS

Just when you thought the browser was the universal client, think again. .
Personally, I don't see RSS functionality being subsumed into browsers. I see it the other way round. Browsers subsumed into other tools - such as news aggregators. I suspect we will see a lot of stuff built around Office 11. Applications that are - first and foremost - Office Apps - but include IE for browsing functionality.
Give away the razors and sell the blades. The blades are Office Apps I suspect (for Microsoft environments.) and IE is the razor.
I think the OpenOffice and Mozilla folks should get into the same room some time:-)

I'm going to buy Paintshop Pro

I've always wanted something like Paintshop Pro but with serious scripting capabilities. I want to do things visually, then automate them. Today on the Daily Python site I find that PaintShop Pro 8 includes Python
scripting. Yummy.

Many years ago we wrapped the Framemaker API in a Python extension and embedded Python scripting into Frame. It was awseome. However the powers that be weren't interested in anything that wasn't Java based. Note to any powers that be - Java is a crappy, crappy language for application scripting. Jython, JRuby, Jackle all run on the JVM and are much better for the task. Repeat after me "Java the programming language is different from Java the platform".

Full text search in Python

[via Don Park]A Python port of the Lucene (Java) engine.

Ogham and Perl regular expressions

[via Walter Higgins]. The similarities between Ogham and the kind of perl RE required for XML.

Code, data and dyadic generalists

In the ITWorld archives Code is data and data is code.

Prelude and fugue in Python arranged for OpenOffice and Strings

OpenOffice is soo cool. python is sooo coool. xml is sooo coool.
This is my current favourite four lines of Python:

import zipfile
import re
z = zipfile.ZipFile ("foo.sxw")
t = z.read ("content.xml")
# And were off!
...


The power of these four lines in the hands of someone who knows a bit about the XML file format used by OpenOffice is significant. Case in point. I wanted to find all the images I had linked into my OO doc. This does it:

images = re.findall ("<draw:image .*?/>",t)
for i in images:
print re.findall ('xlink:href=".*?"',i)

Its not efficient. Its not optimal. Do I give an aspirated fricative? Nope. It works and it took the duration of a sparrows gaseous indescretion to write. Compare and constrast with OLE automation or any other botchery you like. Use the file format Luke :-) Yes, I am having a fight inside my head. I know I should be invoking a WF parser on the content.xml string but gee Ma, I've got work to do.

xml-dev archive has had a facelift

http://lists.xml.org/archives/xml-dev/200303/maillist.html more text per square inch.