My Zeroth Post
Thanks for stumbling across my zeroth post on my very zeroth blog. Why zeroth? We’ll get to that in just a little bit.
Since starting the RUBY-005 class at Flatiron School, we’ve been beginning lecture by focusing on a different programmer each day and discussing their contributions to computing. My favorite so far has definitely been Edsger Dijkstra. Besides him being the father of structured programming (something I am immensely grateful for), I found his sense of humor to be incredibly refreshing. Here are a couple of my favorite quotes of his below:
“About the use of language: it is impossible to sharpen a pencil with a blunt axe. It is equally vain to try to do it with ten blunt axes instead.”
“It is practically impossible to teach good programming to students that have had a prior exposure to BASIC: as potential programmers they are mentally mutilated beyond hope of regeneration.”
Wow, harsh. Pretty much the entirety of my experience with BASIC centered around playing the QBasic game Gorillas when I was in elementary school. That statement alone makes me happy I never took my curiousity further than that.
One of the first hang-ups I had as a coding beginner was the concept of zero-based numbering; that the first element in a given array has an index of 0, not 1. It was really more of a minor annoyance than anything else – everything I had encountered up until this point began with a count of 1, and the habit of thinking in that manner found me making the same mistakes in my code over and over again.
Dijkstra argued in favor of zero-based numbering in a paper he wrote while at the University of Texas. I still can’t say I completely agree with the use of it semantically, but it was very interesting to gain a new perspective on how it makes sense from a mathematical standpoint.
His argument centers around the idea that expressing the range of a
variable i as x <= i < y offers advantages over any other combination of inequalities. Besides his belief that zero is the smallest natural number, I found the following reasons to be the most compelling for its usage:
• In some tasks it’s useful to ask “How many values occured i times?” Placing such values in an array requires beginning the indices at 0; in the case one has the possibility that some values occurred zero times.
• Consecutive ranges have matching end-points: x <= i < y followed by y <= i < z yields x <= i < z.
• The length of the range is the difference of its endpoints: 0 <= i < 10 has ten elements, as does 5 <= i < 15.
• In iterative algorithms that count from zero, the current value of the count always equals the number of previous iterations.
• For a zero-origin array, the index indicates the displacement from the origin.
I believe that the language should adapt to the user when at all possible, not the other way around. However, Dijkstra makes a lot of points that are very difficult to counter. At least for my decidedly un-mathy brain.