Artsy Recycling

Even when you actually want to recycle computer parts (especially scrap parts that do not quite work anymore) it’s quite hard to do so. One possible solution is to simply chuck everything in the usual recycling bin and hope for the best. Or you can try to find a metal reseller. Or you can use the parts in a creative way. Kind of.

I disassembled the CFM01 and got quite a lot of spare parts from the 1U Pentium III servers. The casings aren’t all that interesting since they’re fairly cheap (compared to, say, a Dell PowerEdge server) and the CPUs are useless. Nobody wants them. Even recycling the all-copper heat sink proved a problem. So I used them differently.

Wallpaper: Cement Wave

(Cement Wave, 1920×1200)

Wallpaper: Menaçant tubercule

(Menaçant tubercule, 1920×1200)

Hermite Splines (Interpolation, part III)

In previous posts, we discussed linear and cubic interpolation. So let us continue where we left the last entry: Cubic interpolation does not guaranty that neighboring patches join with the same derivative and that may introduce unwanted artifacts.

Well, the importance of those artifacts may vary; but we seem to be rather sensitive to curves that change too abruptly, or in unexpected ways. One way to ensure that cubic patches meet gracefully is to add the constraints that the derivative should be equal on both side of a joint. Hermite splines do just that.

Asking the Right Question

There are many reasons why you’d like to know if your user liked something or not. It can be used later for recommendation, for example, or to assess whether or not the users want the new feature you’ve just added. And it seems to me that not only the question but also the choices of answers offered matters a lot.

So, let’s say you add an (optional) survey to your application to collect feedback. You can ask a yes/no question like “do you like the new feature?”. The obvious yes/no answer isn’t so obvious. If the users select yes, it means yes, but what if they click no?

Cubic Interpolation (Interpolation, part II)

In a previous entry, we had a look at linear interpolation and concluded that we should prefer some kind of smooth interpolation, maybe a polynomial.

However, we must use a polynomial of sufficient degree so that neighboring patches do not exhibit very different slopes on either side of known points. This pretty much rules out quadratic polynomials, because polynomials of the form are only capable of expressing (strictly) convex (or concave) patches. A quadratic piece-wise function would look something like:

Wallpaper: Rust Matrix

(Rust Matrix, 1920×1200)

More wallpapers can be found here

Wallpaper: Stone Squid

(Stone Squid, 1920×1200)

More wallpapers can be found here

Wallpaper: Rocket to Mars

(Rocket To Mars, 1920×1200)

More wallpapers can be found here

Fast Fibonacci

The Fibonacci numbers are very interesting for many reasons, from seed heads to universal coding (such as Taboo Codes). Just figuring how to compute them efficiently is plenty of fun.

The classical formulation of the Fibonacci numbers is given by the following recurrence: