Thanks to Eric Wise for pointing me to this great article on .NET Design. I just ordered the book from Amazon!

Waterfall 2006 Making Outsourcing Work: One Team Member per Continent by Babu Bhatt LOL!!

Scoble interviews Scott Isaacs!!! (No not our Scott Isaacs). He feels the same way about AJAX as I do. Nothing new except the name.

I have been working on a project recently involving some image processing. The application takes many images of the same thing and attempts to sharpen the image. There are many ways to do this depending on what details of the image are important. One can use simple image averaging (low-pass), speckle imaging and wavelet filters, just to mention a few. This was my first exposure to wavelets and found them to be very interesting. So I thought I'd share a bit here. For those of you more mathematically savvy than me, I'll apologize now. ;) Wavelet transforms are similar in nature to Fourier transforms in that they can take a signal and represent it as a set of coefficients. The primary difference being that wavelets are localized in scale and time where the Fourier transform is only localized in frequency. For those of you that remember Fourier transforms from Calculus, this should be a good indication as to where we are going. There are many kinds of wavelets. One can choose between smooth wavelets, compactly supported wavelets, wavelets with simple mathematical expressions, wavelets with simple associated filters, etc. So how do they work? Lets start by looking at the Haar wavelet. The Haar wavelet is a simple step function. (follow the link to take a look) If we scale (dilate) and move (translate) this simple function we can use a series of them to represent a signal. (pause) OK...so how do I use it to sharpen an image? Well if we take an image, represent it with a set of wavelet coefficients then take another image and do the same, then another and another and so on, what will we see? We will see some of the coefficients hover around a value while others may change wildly. The values with the most constant values will correspond to the most salient features. These features give us the closest representation to the real object in the image. Here's a great example of a C# implementation of a dualtree complex wavelet transform (the one I'm actually using)
Conclusion
I know I've glossed over a lot and over simplified most of it, but I wanted to share wavelets with those of you that have never heard of them. They're very cool!

Thanks to Don Box for pointing me to this AWESOME rap. This is a must listen for all of us!

Just finished reading Manifold: Origin by Stephen Baxter. It was an interesting (fiction) book that is pretty slow at first with 6 seemingly different plots. Baxter covers topics in evolution, space-time, sociology, religion, and so on. It's kindof like Planet of the Apes meets 2001: A Space Odyssey. I give it a 7 out of 10.

Joel posted an article today that discusses hiring the best programmers for a project. I agree with the article but have found that most managers I have worked with agree with the "King David" approach he mentioned. So the trick is how to show a tangable ROI for hiring the best programmers and how do you define the metrics for finding the best progammers?

Here's a quick example of the GOTCHA app I talked about yesterday. It's a class library and test windows form app. It takes a background and as many images as you want. Then it randomly assigns numbers between an assigned Min and Max to each image, and rotates them. I also make sure that there is no image overlap (some could be good though). Finally it generates the question and the answer. The question is in the form "What do you get when you add the {0} and the {1}?"
Where 0 and 1 are the names assigned to the images. Here is a screen shot of the test app:

Let me know if you use this on your web page or blog!

Dan Bartels reports that Microsoft will be naming Longhorn to Windows Vista...Is this what we have to look forward to? :)

Robert Plotkin's just posted an entry on his blog, "Automating Invention," supporting my case about the downturn of software engineering or should I say software manufacturing. It basically states that we are victims of our own ingenuity. By automating so many of our tasks we have opened up the field to Primate Programming.