At 9:53 PM +0900 12/17/02, Austin Ziegler wrote:
>On Tue, 17 Dec 2002 16:49:47 +0900, Rich wrote:
>>  The problem lies in the fact that these statements are equal:
>>
>>  can't be compiled == non-trivial
>
>They're not equivalent, though. Bulat is saying:
>
>   Ruby currently can't be compiled. Ruby needs to become a
>   statically typed language in order to be able to be compiled. The
>   corporate world won't accept Ruby until it is a statically typed
>   language.
>
>This is patently crap, as both Perl and Python have been well
>accepted in the corporate world (although both initially through the
>back-door) and neither of these are compiled or statically typed
>(but neither is as dynamic as Ruby so far as I can tell).
>
>Dan, however, is saying:
>
>   Ruby could be turned into a compiling language without losing its
>   dynamicity -- which provides advantages that statically typed
>   languages can't provide -- but it would take a lot of work to do.
>   It would take a year and four or five scary smart people to work
>   with Matz and do nothing else during that time.

Almost. You could go compiled now, you just won't see the speedups I 
was talking about (the order of magnitude or two one). There's 
nothing about Ruby as it stands now that prevents you from generating 
a standalone executable. (And not just a wrapped interpreter, a real 
executable) Being compiled or not is orthogonal to the sort of 
dynamicity that makes the speed gains difficult.
-- 
                                         Dan

--------------------------------------"it's like this"-------------------
Dan Sugalski                          even samurai
dan / sidhe.org                         have teddy bears and even
                                       teddy bears get drunk