The author is to ignorant of both real programming and real biology to even make the comparison correctly let alone make a significant point.

He make a big deal about life having no legacy code, but (given our current understanding) that is simply wrong. Upwards of half the genetic code may be nothing more then junk code, commented out but not removed.

He trys to handwave away mutation by proclaiming all mutations harmful, trying to draw a connection between conventional computer bugs and mutations. This ignores the fundamental differences between genetic coding and conventional computer coding, and also ignores the fact that there are computer systems designed to evolve code through a mutation/selection cycle.

He trys to draw a parallel between DNA coding and OO computer programming, but computer programming isn't OO in the sense of the word that something like C++ is. If anything, DNA coding is more like Forth or Lisp, where there are a handful of hardwired language components and the rest of the language is built by constructing routines out of older routines and those hardwired basic elements.

His whole argument doesn't make any sense at it's core, and contains numerious gross distortions of fact.