Real men program in C

August 01, 2009

Michael.Barr-August 01, 2009

A couple of months ago, I ate a pleasant lunch with a couple of young entrepreneurs in Baltimore. The two are recent computer science graduates from Johns Hopkins University with a fast-growing consulting business. Their firm specializes in writing software for web-centric databases in a language called Ruby on Rails (a.k.a., "Ruby"). As we discussed many of the similarities and a few of the differences in our respective businesses over lunch, one of the young men made a comment I won't soon forget, "Real men program in C."1

Clever though he is, the young man admitted he wasn't making that quote up on the spot. That "real men program in C" is part of a lingo he and his fellow computer science students developed while categorizing the usefulness of the various programming languages available to them. Exploring a bit, I learned the quiche-like phrase assigns both a high difficulty factor to the C language and a certain age group to C programmers. Put simply, C was too hard for programmers of their generation to bother mastering.

Is C a dead language?
For today's computer science students, learning C is like taking an elective class in Latin. But C is anything but history and not at all a dead language. And C remains the dominant language in the fast growing field of embedded software development. Figure 1 summarizes 13 years of relevant annual survey data collected by the publishers of Embedded Systems Design.

View the full-size image

The discontinuity after 2004 is necessary because the phrasing of the question and permissible answers were changed in 2005. Prior to 2005, the question was phrased, "For your embedded development, which of the following programming languages have you used in the last 12 months?" In 2005, the phrasing became, "My current embedded project is programmed mostly in ____?" Prior to 2005, multiple selections were permitted. This meant that the aggregate data was allowed to sum to over 100% (the average sum was 209%, implying many respondents made two or more selections).

The biggest impact of the survey change from multiple selections to one selection was on the numbers reported for assembly language. Prior to 2005, assembly language was present in an average of 62% of all responses to this question. This should not be surprising, as it is well known that most firmware projects require at least small quantities of assembly code.

After 2004, assembly becomes a minor player--averaging just 7% of all responses across five survey years.2 This data more closely represents the percentage of projects written mostly or entirely in assembly. The data also show a decline in the popularity of that programming style, from 8% in 2005 to 5% in 2009.

< Previous
Page 1 of 3
Next >

Loading comments...