Tim On Computer Science

The title is a “controversial” quote from the book Hackers and Painters. The book’s author calls it controversial; I call it blindingly obvious but still not seen by certain people.

That book serves as an example that, even though programming languages differ, not everyone agrees on the “important” criteria. In his world, Lisp is the best language, so his criteria match that (e.g. recursion, dynamic typing, garbage collection). Lisp’s ability to write a program that creates a program (see “macros”) gets his highest praise, something much more powerful than c’s #define macros. [For my take on the matter, look at my scorn of #defines, below, then imagine making it even more powerful].

Another thing [that I may just be imagining] I noticed was a hint of sour grapes. Specifically, I’m guessing when Yahoo bought his company and its Lisp code, Yahoo immediately began re-writing the system in a language that emphasized other criteria over “power”, for example, “maintainability”, or “readability”.

Finally, his catch phrase “Lisp has no syntax” is just hilarious. It serves as the definitive example of why the book is kind of a test: Can you spot the falsehoods? The book is thought-provoking, sure, but falsehoods abound nonetheless, sometimes in a Jabberwocky “words mean what I choose them to mean” way. My reference on macros, above, must drive him nuts since the first line is “Macros in Lisp provide a very powerful and flexible method of extending Lisp syntax”.

I just ran into the phrase “commoditization of intelligence” from Accelerando (Singularity). In the book’s context it is scary enough. But it is truly scary when it manifests itself today. Like when knowledge workers are treated like replaceable cogs in a software factory. Or when programming languages are lumped together with “it just doesn’t matter”.

Some technology has almost been commoditized – like the electric outlets in my house. Although, “almost” in this case means “you have heard of GFI, surge protection and true-sine wave UPS, right?”. “Commoditization” in this sense implies a lack of knowlege, a lack of intelligence to distinguish trivial differences from the important differences. Hence, the shock (irony?) of the phrase “commoditization of intelligence”.

I’ll continue building my own computers, because I find that even with commodity pieces, the complete solution provides an order of magnitude boost in the performance-for-price ratio.

I just read about Scala for a few hours. Mainly from Java to Scala with the Help of Experts under the sections “Scala for Java Refugees” and “The busy Java developer’s guide to Scala”.

My first take-away is “Good Luck With That”. Where “That” is code like this: (based on Listing 9)

def main(args : Array[String])
  {
    tryToFigureThisOut     // "This is not part of the language"
    {
      thisMethodThrowsExceptions_GoodLuck
    }
  }

It reminds me of this [invalid but representational] c code:

#define while(j,k) printf(j,k); fflush(stdout);
#define F for(r=d;r<n+N;
F++r)*r=c;
while (--y, --x);

My second take-away is that Scala is a nice spoiler for Ruby – I can see people who are drawn to the wild world of Ruby settling for Scala instead. It is crazy enough to scratch the itch, but backwards-compatible.

Welcome to Tim On Computer Science.  This site is dedicated to various thoughts and comments on topics concerning Computer Science.

Continue reading →