Recently while reviewing other folks methodology for "comparing" programming languages, I resolved that the common practice of searching for the number of 'hits' in a web search engine was a pretty poor (yet seemingly standard) measurement.

In response, I resolved to a different variant of this measurement standard: Instead of searching for the language name, search for "HelloWorld.foo" where "foo" represents the common 'extension' for that particular language.

This alternate methodology is more desirable because it does not compare the 'total number of hits' instead, it allows one to evaluate the *content* of the hits, and thereby assess some of the fundamental characteristics of the language in comparison to others. For example, one can compare how many 'hits' consist of:

- various complaints that HelloWorld.foo will not run 
- tutorials for new learners
- a part of online documentation (if it exists)
- discussion about programming in general
- a part of 'social banter' (public mindshare)

[download]

The HelloWorld metric appears to have quite a bit of promise for getting 'big picture' comparisons. A cursory initial review of the "HelloWorld.pl" metric seems to indicate that most of the "complaints" type posts are not even related to perl, but to CGI and web programming.

Comments and critique welcome.

Replies are listed 'Best First'.
Re: Programming Language Comparison Level 0: the HelloWorld metric
by hardburn (Abbot) on Nov 02, 2004 at 14:22 UTC

    I have reaffirmed a long-standing and strongly held view: Language comparisons are rarely meaningful and even less often fair. A good comparison of major programming languages requires more effort than most people are willing to spend, experience in a wide range of application areas, a rigid maintenance of a detached and impartial point of view, and a sense of fairness. -- Bjarne Stroustrup, The Design and Evolution of C++

    "There is no shame in being self-taught, only in not trying to learn in the first place." -- Atrus, Myst: The Book of D'ni.

[reply]
Re: Programming Language Comparison Level 0: the HelloWorld metric
by TedPride (Priest) on Nov 02, 2004 at 15:08 UTC
    Every language has something it's good at, so comparing language x with language y is pointless unless you define specifically what you want to use the language for:

    - I use Perl for algorithms / site maintenance
    - I use PHP for embedding variables and simple routines in pages
    - I use C++ in those rare times when Perl isn't fast enough
    - I use Javascript for client side stuff (focusing a form field after the page loads, popping up an alert if someone forgot to fill out a form field, etc)

    Why try to compare apples and oranges?

[reply]

      Good point, however please note that the OP doesn't say anything about *ranking* programming languages, but just *comparing* them.

      It may seem like a subtle difference, but the point is that comparison is *not* pointless when you want to: 1) compare two or more languages that are potentially capable of mutual substitution (eg deciding b/t Java and CSharp dotNet, SVG or Flash); or 2) get "big picture" aspects of a language you are not familiar with (eg how difficult is it to install, find documentation, etc); or 3) discern its principle proponents and industry niche (eg, finance, law, medicine, manufacturing)

      The very fact that people tend to equate "comparison" with "choosing which is better than the other" underscores my point that a non-emotion-laden metric like the HelloWorld.foo test has some merit.

[reply]
Re: Programming Language Comparison Level 0: the HelloWorld metric
by neniro (Priest) on Nov 02, 2004 at 15:20 UTC
    If you like comparisions between popular scripting-languages take a look at the scriptometer.
[reply]

Back to Meditations