Andrew Que Sites list Photos
Projects Contact
Walnut, Maple, Zebrawood, and Mahogany

Walnut, Maple, Zebrawood, and Mahogany

   The most lovely piece of lumber I have ever bought.  This is Zebrawood.  I spent a long time looking through the supplier's selection of Zebrawood before selecting this piece.  Very vivid grain patterns.  I decided to wide it down with mineral spirits to get an idea of how it will finish.  It is even darker than black walnut, and I will have to take this into consideration for my project design.  Isn't it lovely though?

May 05, 2014

Polynomial Regression Online Tool Compleat

Julia and Ben at breakfast

Julia and Ben at breakfast

   After a fair bit of work my Online Polynomial Regression Calculator is complete.  The JavaScript/PHP interface allows regression analysis of data.  This was fairly easy to implement because the polynomial regression library has already been written.  Most of the time spent on the project was to make a user-interface that was intuitive and could handle various formats of input data.  I played a bit with CSS as well to try and make the interface look nice.
   I am not sure how much demand there will be for such an interface, but I have seen search requests for online polynomial regression.  So this should satisfy those requests.
   My Polynomial Regression Class uses arbitrary precision arithmetic, which is more CPU intensive than conventional floating-point arithmetic.  As such, I had to limit the interface to a maximum of 1,000 data points, and no greater than 9 coefficients.  I figure anyone wanting to analyze more data than this, or needing more coefficients can either find an other tool to work with or download the libraries and run the analysis on their own machine. 
   In working with the CSS I discovered that input check boxes are rather complicated.  The border settings that work for other inputs do not apply to check boxes, and one needs to draw a graphic if they are to be customized.  So I started with a template I found online and drew my own just for practice.  It looks good and works like I want, but I was hopping that with all the crazy stuff that can be done with CSS3 check boxes would have had a little more configuration available.  Should I want to change the theme of this page, I would have to go back and change the graphics.  Maybe in the future they will add more for check boxes.

May 02, 2014

Apache 2.4 still not playing nice.

Young drummer, and not too bad.

Young drummer, and not too bad.

   So after a lot of playing around and getting comfortable with Apache 2.4 it was time to build and install it on the server.  I wanted to get the setup working with the latest version Apache, PHP, and OpenSSL.  After getting the setup to compile, I installed Apache to my home directory on the server.  This allowed me to import my configuration script and update it.  I set it up to run on secondary ports: 8080 for HTTP traffic, and 8443 for HTTPS.  This allowed me to test the settings and verify all the services and virtual hosts were functioning.
   After everything seemed like it was ready for the switch, I changed the test server configuration to the primary server ports.  Then I shutdown the normal Apache server and tried to start the test server.  I kept running into the error "Illegal attempt to re-initialise SSL for server" and Apache 2.4 refused to start.  It only effects the HTTPS part of the configuration.  If I change the listen port for HTTPS to 8443, it works fine.  For some reason it refuses to work when configured for port 443.  Searching around for this came up with nothing.  So for now the switch is going to have to wait.

May 01, 2014

Stats for April

So the number for April are now in. We received 10,692 visits (4,797 unique). This is down from March when we had 15,225 but up from February (9,807) and January (9,932). In 2013 our site ranged from just over 9,000 visits to over 24,000 visits a month, although I haven't been able to deduce what is responsible for the traffic fluctuations. More than half all the traffic to our site are bots, and it is unknown what the split is between legitimate and malicious.

May will have a good deal more filtering. The ratrap will eliminate much of the malicious bot traffic, and I can use the logs to discover what areas of the site are actually popular.