Thursday, January 23, 2003

Frodo has failed ... click on this link to find out how and why. :-D

Lesson in life #3 : DRIVING FAST DOESN'T GET YOU THERE ANY FASTER , IT JUST STRESSES OUT YOUR ENGINE !
Slow the f*** down and enjoy life.

It's that time of day, once again ... to turn inward and say "f*** you" to all those who would oppose my "inner introvert."

Life has been quite interesting during these past few days, while I've been attempting to adjust to my new class schedule. The homework actually looks fun and exciting, no sarcasm intended, and the new student union building looks awesome ... certain parts of it remind me of a dream I had a while back, of a place with funky architecture. (The dream changed from nice to nightmare to nice, but I still love the new addition to the school.)

I've been learning quite a few interesting bits of knowledge about programming, and have learned sickening things about the often hyped ideas within the computer science industry ...
For instance, object oriented code is actually an ugly thing as far as compiling it is concerned ... I had always assumed that the reason for its use was not simply because it made source code easy to maintain.
Another interesting fact: reference counting-style memory management sucks.
Also, Garbage collection, if done correctly, is often as good as explicit memory management done by programmers ... although my own interpretations of the results from such papers as Zorn's 1992 comparison of 6 memory management algorithms was that the programmer's optimizations were very useful in many cases. Still, as shown here, garbage collection is quite the contender. After tweaking around with the weighting of the scores of the different language implementations, it became apparent to me that OCaML narrowly beat out the C programming language in The Great Language Shootout: after near-equal scoring in memory and cpu usage, the deciding factor was the size of the source code. I was suprised ... OCaML is object-oriented, type-inferring, and garbage-collected, if I'm not mistaken. That's fairly impressive, so there is some hope for my ideas. Later on, I'll take a step back and look at what simple language features keep the performance cost low in assembly and MSIL. In the meantime, I'm going to learn OCaML, and see how the progress of the creation of OCaML.NET goes!

No comments: