6 thoughts on “5754463f

  1. Also sorely lacking is Stevens “UNIX Network Programming”, TCP/IP Illustrated, etc.

  2. Actually I would not rate Knuth’s Art of computer programming in the top 100. It is very useful for cribbing citations from but that is all. I don’t share the author’s enthusiasm for machine code description of algorithms.
    I would go further in fact, the ACP approach is part of the problem, not the solution. There was much more to programming than just algorithms even when Knuth started writing.
    The idea that we should encourage the view that computing should be an arcane profession centered on hard to explain and understand ‘folklore’ continues to be a major problem in the field. I suspect that many of the people who prefer MIX do so precisely because it makes the knowledge less accessible.
    The ACP manages to make the source material even less accessible than the original papers in which it was published. There are much better compendiums of algorithms available. This is not a book I would recommend that anyone read, much less copy.

  3. I admire Donald Knuth more as a mathematician than as a programmer. His programming style dates to a time when computer memory was a few thousand words; I think clarity is now more important than size. But volumes 1 and 2 of TAOCP were the best introduction to discrete mathematics that I had ever seen, and they influenced me profoundly when I read them.

Comments are closed.