CL proven to be the best programming language ;-)
Monday, June 14, 2004
The Great Computer Language Shootout test has been
revived. Basically, this is a page that was designed to test all the
major scripting languages by comparing results of a number of
different types of tests. The latest version of the test has been
updated with new languages and revised to work with modern compilers. Although the authors admit that the results
should not be taken as any meaningful representation of the relative
capabilities of different languages, computer language statisticians
everywhere will have a field day using these tests to prove that
their favorite language is the "best".
There is
a page that allows you to assign different "weights" to either
specific tests or to the 3 main areas that are tested:
- I used the default weightings for each of the individual tests.
- I assigned a multiplier of "5" to the CPU time score as CPU speed is normally the primary factor that people compare when evaluating different languages.
- I assigned a multiplier of "0" to the Memory score as the testers state that: "this sampling method has proven unreliable"
- I assigned a multiplier of "1" to the Lines of Code metric as LOC should be considered important in measuring the ability of a programming language to represent higher levels of abstraction.
Before you flame me on this posting, just remember this old adage:
"There are lies, damn lies, and benchmarks."UPDATE: I wrote this entry on Sunday and, at that time, the CRAPS calculator was showing CL as the top programming language based on the criteria I specified above. However, on Monday morning, prior to posting this entry, I had another look at the page and saw that CL had moved from first place to fourth place using the same criteria. Apparently, either the metrics code or individual sample results had been changed overnight. Oh well, fame is fleeting. :-)
