begin quoting Andrew Lentvorski as of Fri, Feb 15, 2008 at 10:00:43PM -0800: > Mark Schoonover wrote: > >Well, it's going slow - atleast for me. Partially because of outside > >influences. I'm currently working through section 2.3.2, but I'll tell you, > >Lisp is starting to drive me a bit batty. Maybe I should stop SICP, and > >actually learn Lisp. > > Then I recommend "The Little Schemer" and "The Seasoned Schemer". They > look silly, but just do the exercises. They tend to be in byte-sized > chunks that you can do at your own speed.
I should add those to my "want" list... you keep mentioning 'em. [chop] > > I'm thinking I could be fighting previously learned > >habits from learning assembly, Pascal, C and Perl that Lisp is sufficiently > >foreign enough that I'm not able to understand what the code is telling me. I discovered that it took about two hours to twist my brain into a shape where scheme-think was natural, and that my brain wouldn't stick in that shape unless I did some scheme every day. > It is not just you. Many people complain about the impedance mismatch > between Lisp and algorithms. Lisp maps to recursive/tree algorithsm > very well. Iterative and decision algorithms, not so much. Mergesort in Scheme is a dream. > Of course, doing tree-based algorithms in Java and C tends to be a > painful experience. Um.... why? And do you have an example? All I can think of is that in C (and Java), you have to define your tree data structure, or you have to use a clunky generic predefined one. It's not until I have to play with XML that I start feeling pain with tree-structures in Java or C, but I suspect that's a library misdesign problem. -- I really do like the idea of code blocks "Here, do this to every one of those!" Stewart Stremler -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg
