--- In [email protected], Thomas Hruska <[EMAIL PROTECTED]> wrote: > > Brett W. McCoy wrote: > > On 8/26/07, Nico Heinze <[EMAIL PROTECTED]> wrote: > > > >> To anyone who's serious about writing good code: > >> IMO the best basis to programming in general is > >> still provided by Donald Knuth's The Art Of > >> Computer Programming; recently I received the three > >> books as a birthday gift, and you should see my > >> eyes flashing brightly when I'm sitting on board > >> an airplane (every Monday I'm travelling to my > >> current project site, every Friday I'm flying back) > >> and read Knuth's classical work. I'm still in the > >> preparative chapters, but even though I am by far > >> not good enough with mathematics to understand > >> everything Knuth has written I simply enjoy every > >> line I read and every exercise I try. This is the > >> best introduction into serious programming I've > >> ever seen. I only can recommend working through it, > >> and I well know that not many students ever touched > >> this book even if they had the chance. > > > > I only have the first volume right now... he's a good > > writer, no doubt about that, but the material is still > > pretty difficult, at least to someone who has no > > formal CompSci education. > > > > -- Brett > > Hmm...I've never read those books. Probably because I've > heard they are huge and difficult and I like reading books > cover-to-cover in a single sitting (the main reason why > I'll probably never get around to reading War and Peace > either).
Under this assumption you probably never will touch The Art Of Computer Programming. ;-) Honestly, Knuth writes in the preface that this is not meant to be read cover to cover. The reader is supposed to sit down and work through exercises carefully; even if she fails, he shall at least try hard and honestly. > Including technical manuals. Yeah, I'm weird. > But we already knew that. Not any more weird than I am: Many years ago I was lucky enough to get hold of a book about COFF (Common Object File Format, a file format defined by AT&T about compiled object code [for those who didn't know this abbreviation]) which is supposed to be tiring even to hardcore-insomniacs. And I even enjoyed reading it, at least until I was halfway through. You see, you are not the only weird one. ;-) > The math sounds interesting. What level of math > background do you need? Actually he's introducing everything he needs in the latest revision of the book. He's extended the math introduction considerably since the second revision. Yet it's been many years since I've last worked with matrices, determinants, and the like, so this whole chapter is almost like a book with seven seals to me (a German expression meaning that it's almost a complete mystery), yet it's plain fascinating to read anyway. You have to be familiar with the basic math operations; with integer, rational, and real numbers (though they are introduced in some detail); but I'm sure you won't find it too hard. BTW Knuth explicitly mentions in the preface that many of the "exercises" in the books are extremely hard to do; he rates all exercises from 0 to 50, 0 being extremely easy and 50 being extremely hard. Personally I don't agree with his rating of Fermat's last theorem (the thingy regarding a power n + b power n + c power n having no positive integral solutions for n > 2), he rates it as 45, I think this is too low. But in the end the mathematical stuff is interesting to me but by (almost) no means mandatory in order to understand the remainder of the book. Through most of the mathematical stuff I just skim because it's too high-level for me. Yet I know that I'll get the gist of almost all algorithms he presents throughout the book without trouble; I've read them often enough to know this. Lucky me. :-) Regards, Nico
