From: Garance A Drosihn Date: 2003-08-21T19:39:52+09:00 Subject: Re: Punctuation as noise >Hal E. Fulton wrote: >> >> As a very crude way of measuring this, I decided >> to count alphanumerics vs. symbols in code >> (counting whitespace as neither -- an arbitrary > > decision on my part). I understand what idea you're getting at, but I think this script is too crude. It can be completely thrown off if one script happens to use longer variable-names than another, for instance. I tend to use longer names than the average C programmer does. Also, when I'm writing a script, the longer the script gets, the longer my variable names are likely to get. One possible side-effect: If it takes me more lines of code to implement a given task one language than another language, then the "longer" language is likely to have longer variable names, and thus score lower (aka "better") with this punctuation-scoring method. The punctuation-scoring method would also favor languages which use a lot of long keywords. I once worked with a language which used long keywords by design. All reserved words were more than letters, because it assumed that programmers would prefer to pick shorter variable names. So, it did things like use 'whenever' instead of 'if'. It also favors languages which force you to use words for what most languages use punctuation for. I wonder what kind of a score cobol would get using this method. [at the other end of the scale, APL would not fare well with this method, and for very very good reasons! :-) ] So while this starts out sounding like an interesting measurement, I think that as soon as you try to get "hard numbers" (which look significant simply because they are "objective values"), you'll find that the measurement is pretty close to meaningless. -- Garance Alistair Drosehn = gad@gilead.netel.rpi.edu Senior Systems Programmer or gad@freebsd.org Rensselaer Polytechnic Institute or drosih@rpi.edu