So Gmail added IMAP support this month and the biggest benefit I got from the new addition was a simple way to import my pre-Gmail email. When I first got my account several years ago I looked into importing all of my old email messages into my Gmail account but found it would be tedious and fraught with problems. The biggest one being there seemed to be no way to import the emails and preserve their old header information such as who sent it, what address they sent it to, when they sent it and when I originally received it. All the import tools I found were nothing more than glorified resenders that simply emailed my old emails to my new account. But now with an IMAP link between Gmail and Apple Mail, my current email client, importing my old email is as easy as dragging and dropping them into my Gmail inbox. Unfortunately Apple Mail 3.0 is still flaky and trying to move too many emails at once causes it to crash. If I import them in about one-month batches it seems to handle it...

I recently was forwarded an email about Google's "scary" (for some) feature of reverse phone searching. You can read more about the email at Snopes.com . Anyway, I was bothered enough by the false sense of security a FUD email like this provides for those who aren't as wise to today's Internet issues that I was forced to write the following response for my family to read as a counterpoint. I receive many forwarded emails, but I award few with more than a passing glance. This one is so misguided though, I felt it deserved a response at least to those I know and care about. Phone numbers and addresses aren't private information and shouldn't be or they betray the very nature of what they are, location devices. Reverse searching for a person or address based on a phone number has been available for a very long time on many websites, the most well known, for me at least, being http://anywho.com. The advice the original email provided will lead many who rea...

So try this in Maven. Create a project A and don't install it. Create two subprojects, B and C, that have A as their parent. Make B depend on C. Install C. Try to build B. You won't be able to as Maven claims it can't download A. Huh? Why did C build then? The cause is that Maven can resolve a project's parent on the local drive using a relative path. But if a project has a dependency, it will only look up that dependency's parent through the repositories, even if that parent is available via a relative path. The solution, install A into your local repository. If another developer gets B out of your SCM and tries to build they will fail with the same problem. So I've now learned. Deploy more snapshot builds of parent POMs to keep my fellow developers happy.

My coworker asked me to describe how Maven compares and differs to Ant. I realized how hard it was for me to describe what a developer gets for moving their build process from Ant scripts to Maven. I can't make the argument that you can do X with Maven but not with Ant because I don't really have a valid example. However, I am inclined to make the argument that you can do X with both but with Maven it's easier. With Maven, you collect metadata about your project instead of writing scripts defining the steps your build will take. For example, if you look at an Ant script you'll immediately see that it is organized as targets which many developers use to define and group the steps of their build process. In a Maven POM file, you won't see similar build steps defined. This can be confusing for a new Maven user coming from the world of Ant. I know it was for me. Instead, you'll see lots of metadata about the project such as dependencies, Maven plugins to use...

Jen, had a report on The Weather Channel! It's the first time I got to see her on TV. I took the time to grab it off my TiVo, decrypt the video, reencode it, trim out the leader and trailer and upload it to YouTube. Take a look!

So I was brushing up on my CS fundamentals recently and was reminded of the topic of Markov chains . In college, I created an applet that simulated flocking behavior in animals using a Markov chain and it turns out its still up on my college site.