I finally posted the résumé I have been working on.

I fight back against telemarketers who abuse my phone, and now I’m fighting back against robots who abuse my web site. (1871 words, 97 comments) [dive into mark]

This is a great article about slowing down spam by stopping selected spambots, spybots, and unwanted robots. I am surprised I missed it when it was published in February.

Step Away From the Spell-Checker. A University of Pittsburgh study reveals that using the spell-check button can create problems for writers and editors who rely too heavily on the software. While a spell-checker can identify some misspelled words, it is not always so reliable when used to correct grammatical errors. [Wired News]

Actually this problem has existed for some time in different forms. Way, way back in time when I was going to school the same concern was expressed over calculators. At the time the concern were expressed that the general population would gradually cease to be able to add, subtract, multiple, and divide without the benefit of a calculator. I don't see where people's computation skills have been affected. Most people are too lazy to find their calculator. The challenge has always been to avoid complacency and laziness when using a tool like spell-check. Spell-check requires a higher mental effort since you are looking at two more refined and sophisticated groups of errors, errors that you have made and spell-check does not catch and those errors that spell-check finds but are not actually errors.

I think I finally figured out why java liked the windows path format on the classpath option. Even though I am running shell script under cygwin when I crank up java it is running the version that I installed under Windows.

I finally got the makefile that drives résumélibrary to work completely. This really was a waste of time but I was curious. This java stuff had me going for awhile. Actually the big problem was with the path used in the classpath parameter when invoking java. It couldn't find the class to run it. I think I finally settled on the windows version of the path. My logic says it should have looked like the unix path since I am running it under cygwin but I was wrong. I ended up fixing up two shell scripts and having the makefile call them rather than running java directly as was written. I also modified the makefile to make the rtf in addition to the html, text, and pdf. The nice gui app the folks at xmlmind created to convert docbook to rtf ran without an obvious error message. However, the rtf file that was output was unchanged from the input. Oops! I got the cmdline version to run so I should be able to make the gui work, too.

I finally got my résumé updated to work with the latest dtd from resumelibrary. I ended up installing cygwin, Xerces/Xalan, and modifying the classpath. I got conflicting recommendations on classpath. I ended up setting an environmental variable that pointed to the directory with the jar files. Sun thought I should enter the -classpath option on the command line. I tried both and the environmental variable was easier. Xalan immediately pointed out some dtd related errors and offered to upgrade my xml to the new version. The error messages are relatively good so far. I liked Mark Pilgrim's xsl but I can't use it with the new dtd. The new xsl is difficult to figure how to customize. On the whole, this exercise has been fruitful. I have successfully worked through an updated dtd, worked with two xml parsers, and improved my résumé quality and content.

I got nailed yesterday with resumelibrary. I was working from a copy that Mark Pilgrim had cobbled together and then I thought I would get fancy and use the latest dtd. Bad idea! I blew a lot of time trying to figure out why I couldn't get my resume in xml to validate. It seemed so close but I was really so far away. I finally decided that it was interesting enough that I should commit some time so I am going to get my resume to validate with the latest dtd. The first problem to fix is the lack of error messages by the parser. The online version of resumelibrary was not much help. I will try running the parser, xerces/xalan, locally to see if I can get a better location of the problem. I will also give quick a check to saxon and the M$ parser for kicks.

Paul Valery. “A poem is never finished, only abandoned.” [Quotes of the Day]

I started work on my resume today and ended up installing Python, wxPython, and Saxon. Actually I installed Instant-Saxon as I gradually figured out I didn't need the full Saxon. I was looking Mark Pilgrim's resume in XML format and figured out that he had used resume library. So I after downloaded the latest copy off of Sourceforge, I started to download Xerces/Xalan. I was about 25% done when I noticed that Mark had used Saxon and Python. A little internet search later and I had identified that Saxon was a much smaller program and Instant-Saxon was really small. For kicks I tried Instant-Saxon on Mark's xml resume and it worked. Mark really didn't need to use Python. He uses it to create a Word document and there are easier ways than using Python.

The President—whoever he is—has to decide. He can't pass the buck to anybody. No one else can do the deciding for him. That's his job.
Harry S. Truman

Tonight President Bush sounded a lot like Harry Truman. During his speech tonight he said we are going to call for a vote to disarm Iraq with force even if we don't have the votes. Twelve years of passing the buck on what to do about Iraq's infractions is going to stop. If UN is to exist in the future, France, Germany, and Russia will either vote to enforce the resolution or get out of the way. Maybe then Saddam will decide to flee and avert a war.