Hi,

In message "[ruby-talk:15168] Re: Discussion on new Ruby features"
    on 01/05/16, Christian Szegedy <szegedy / or.uni-bonn.de> writes:

|Basically, I don't think that too much optimization is possible without
|giving the compiler some hints...

Agreed.  Ruby really is a hard-to-optimize language.

|My hope is that I can get into some fruitful discussion with Matz on
|this topic. 

I'm interested in *optional* argument type specifier as in Guy
Decoux's enhancement, not for optimization, but for usability of the
language.  It would NOT be in the language in the near future, but
might be in someday.

As for optimization, I think I'm going to focus on optimizing
implementaion.  Yes, I'm a programmer just as much as a language
designer.  From my prototype result, Ruby interpreter can run several
times (more than 10x times in some cases) faster.  Thinking of growing
CPU power, it would be enough for most of the cases.  And when Ruby
would be worth it, Lisp people would help us to boost up.

							matz.