Reputation
13,937
Next privilege 15,000 Rep.
Protect questions
Badges
5 31 49
Newest
 Nice Answer
Impact
~529k people reached

Jan
17
revised Will we go back(?) to fixed-point arithmetic in the near future?
Clarify cost of hardware integer divide vs. hardware floating divide. Add historical note.
Jan
17
comment Why do we still use floats?
The dedicated devices frequently have to do a LOT more processing, in a limited amount of time, than the "general computing platforms". The dedicated devices generally have far lower power and cooling budgets. Last time I looked, you couldn't put an Intel flagship processor in a pocket cellphone, because of the power and cooling requirements.
Jan
17
comment Will we go back(?) to fixed-point arithmetic in the near future?
@delnan: The key concept is that floating-point operations (all of them, not just division) are easier FOR THE PROGRAMMER, and not much harder for the machine. Fixed-point integer computation requires the programmer to do a huge amount of MANUAL bookkeeping to keep track of the decimal (binary) point. This is the kind of thing that computers (at first, software libraries) and hardware (later) are very good at. Read Hamming's book.
Jan
17
comment Will we go back(?) to fixed-point arithmetic in the near future?
@delnan: Yes, of course. However, the results then have to be renormalized, and final exponents calculated.
Jan
17
comment Why do we still use floats?
Actually, a lot of audio signal processing is done with integer DSPs. For speech, 16-bit sampling and integer processing is quite sufficient. For music, 24 bits is probably preferable: the Access Virus synthesizers use Freescale 24-bit integer DSPs. 32-bit float DSP doesn't buy you anything over 24-bit integer DSP.
Jan
17
answered Will we go back(?) to fixed-point arithmetic in the near future?
Jan
14
awarded  Nice Answer
Jan
14
answered Why do so many languages treat numbers starting with 0 as octal?
Jan
13
revised What should be the maximum length of a function?
Correct capitalization of FORTH.
Jan
9
comment Is OCaml any good for numerical analysis?
MATLAB. Numerical solution of systems of differential equations is a solved problem, and has been for quite some time. If your objective is to crunch your numbers, use the tools that have been developed over the last several decades to do just that. Don't reinvent the wheel, no matter how much fun it might be. You will save yourself a lot of time and aggravation.
Jan
1
comment What should be the maximum length of a function?
@CaffGeek, time to comprehend is pretty strongly correlated with raw SLOC. Once the programmer has to "turn the page" (flip a printer page, scroll a screen, look at a different file), his comprehension goes down DRAMATICALLY. Read Weinberg's "Psychology of Computer Programming" for more information.
Dec
31
awarded  Good Answer
Dec
31
revised What should be the maximum length of a function?
typo
Dec
26
comment Why it is `(cons 1 (cons 2 (cons 3 nil)))` and not `(cons 3 (cons 2 (cons 1 nil)))` for [1,2,3]?
Have you read McCarthy's 1960 paper, "Recursive Functions of Symbolic Expressions and Their Computation by Machine, Part I"?
Dec
23
answered Using a “spherical frustum” for 3D projection
Dec
23
answered If mutual exclusion is not implemented, how would we detect a race condition?
Dec
19
comment How to model an arbitrary number of types
It sounds to me as though you're afraid of your compiler. If you can partition your code, so that the behavior of your nucleotides is nicely encapsulated, then there's no reason not to recompile the capsule and link it with the precompiled rest of your program. At that point, you can use a SMALL program to prepare the guts of the capsule, working from a mini-language. LISP excels at this kind of things, but you can do it in other languages as well.
Dec
10
answered When was source control invented?
Dec
9
comment Does a GPL Bison grammar infect my application?
Bison generates parsers for LALR(1) languages. If your language is inherently LALR(1), there are probably some fairly serious crocks in it that make it difficult to USE as well as difficult to parse. You will probably get a lot more lift from simplifying your language enough that you can write a straightforward recursive descent parser (RDP) for it. RDPs tend to be significantly easier to write and maintain than LALR(1) grammars and parsing actions.
Nov
26
answered How strict should you be about indentation/white space?