3 ms·

I often find myself wishing for a combination of the two languages. An S-Expression based Haskell-alike instead of an S-Expression based Python-alike (yes, yes

I often find myself wishing for a combination of the two languages. An S-Expression based Haskell-alike instead of an S-Expression based Python-alike (yes, yes I know - Lisp came first - whatever). Bonus points if I can drop into an S-Expression based Fortran-alike when bashing on a chunk of memory is the only way to do something fast. I like macros, but, as a scientific programmer, I know the type of everything in my program (double-precision floating point numbers and aggregations thereof), and I don't want to play some guessing game where I figure out which types I have to declare to get something to run fast. Same with garbage collection (what! the GC might move this while LAPACK bashes on it? I need to explicitly tell it not to? Fuck!). emacs + paredit is also the best structured editing tool I have used in any language.

kanak 16y ago
Racket has a lazy module that can let you have lisp+laziness. There's also typed scheme which lets you do lisp+types. Now if only someone could make a lazy typed racket.
- neutronicus 16y ago
  Yeah, I've been more of a racketeer than a lisper of late. I like their 'batteries included' philosophy. I don't much like the macro system, though. I might check out typed racket. I don't care too much about lazy evaluation one way or the other, long as all my doubles get multiplied, added, and divided, and it happens fast.
  - astrange 16y ago
    Well, just as general advice, the fastest way to multiply your doubles is to use single-precision SIMD. Does it do that?
    - jedbrown 16y ago
      Huh? Double precision SIMD? SSE2 instructions aren't slower than SSE.
      - astrange 16y ago
        No, single precision, as in "the fastest way to multiply a double is to use floats instead". I've seen more than enough C programs unnecessarily using double that I wish floating point constants would just be float by default. (Although it seems like clang does something special here, as I don't see many double-to-float conversions in its output.)
        jedbrown 16y ago
        Then you aren't multiplying doubles, it would be a shame to get a completely wrong answer on an ill-conditioned problem because the implementation disregarded the spec, cf. "the fastest way to process UTF-8 is to truncate it to ASCII", "the fastest RSA encryption is ROT-13". Also note that floats are only faster with packed instructions (often not possible without changing APIs) or when memory bandwidth is an issue.

kanak 16y ago

Racket has a lazy module that can let you have lisp+laziness. There's also typed scheme which lets you do lisp+types. Now if only someone could make a lazy typed racket.

neutronicus 16y ago
Yeah, I've been more of a racketeer than a lisper of late. I like their 'batteries included' philosophy. I don't much like the macro system, though. I might check out typed racket. I don't care too much about lazy evaluation one way or the other, long as all my doubles get multiplied, added, and divided, and it happens fast.
- astrange 16y ago
  Well, just as general advice, the fastest way to multiply your doubles is to use single-precision SIMD. Does it do that?
  - jedbrown 16y ago
    Huh? Double precision SIMD? SSE2 instructions aren't slower than SSE.
    - astrange 16y ago
      No, single precision, as in "the fastest way to multiply a double is to use floats instead". I've seen more than enough C programs unnecessarily using double that I wish floating point constants would just be float by default. (Although it seems like clang does something special here, as I don't see many double-to-float conversions in its output.)
      - jedbrown 16y ago
        Then you aren't multiplying doubles, it would be a shame to get a completely wrong answer on an ill-conditioned problem because the implementation disregarded the spec, cf. "the fastest way to process UTF-8 is to truncate it to ASCII", "the fastest RSA encryption is ROT-13". Also note that floats are only faster with packed instructions (often not possible without changing APIs) or when memory bandwidth is an issue.