>I think now I understand what allowing Unicode in a programming language >might bring. > >Watch the talk of Guy Steel > http://www.infoq.com/presentations/fortress-steele >in particular around the following times: > >16:00 min -- 19:35 min >34:00 min -- 42:00 min >52:00 min > >and maybe the whole talk might be interesting for some of you. > >Happy streaming... > >Ralf
His description of generators/reducers is very close to the Google map/reduce algorithm. See <http://labs.google.com/papers/mapreduce.html> I do like his regions/distributions idea but he only seems to take this to the core level of a processor. A pentium or gpu has parallelism in the instruction set (e.g. psubb) which does a SIMD parallel subtract within a single object. And the Sage group spent a lot of time working on the slicing of matrices to optimize the algorithm for cache lines. I suppose one could rewrite the library implementation. I do like the unicode extension idea for operators. I would like to pick up the standard lisp ideas of (drop 2 (1 2 3 4 5)) -> (3 4 5) (take 9 (cycle (1 2 3 4))) -> (1 2 3 4 1 2 3 4 1) (interleave (:a :b :c :d) (1 2 3 4)) -> (:a 1 :b 2 :c 3 :d 4) (partition (1 2 3 4 5 6 7 8 9)) -> ((1 2 3) (4 5 6) (7 8 9)) (map vector (:a :b :c :d) (1 2 3 4)) -> (<:a 1> <:b 2> <:c 3> <:d 4>) (apply str (interpose \, "abcd")) -> "a,b,c,d" (reduce + (range 100)) -> 4950 with a larger set of symbols I can easily make these look like standard mathematical notations. And, of course, the latex-like output as an input form would interact very well with a literate description of the algorithms. Tim _______________________________________________ Axiom-developer mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/axiom-developer
