logoalt Hacker News

jerfyesterday at 3:56 PM1 replyview on HN

I also add the observation that while the dynamic typing languages are all growing in the direction of the statically-typed languages, no statically-typed language (that I know of) is adding a lot of dynamically-typed features. If anything the static languages trend towards more static typing. Which doesn't mean the optimum is necessarily "100% in the direction of static typing", the costs of more static typing do eventually overwhelm the benefits by almost any standard, but the trend is universal and fairly clear.

I kind of think there's room for a new dynamically-typed language that is designed around being fast to execute and doesn't cost such a huge performance multiple right off the top, and starts from day 1 to be multi-thread capable, but on the whole the trend is clearly in the direction of static typing.


Replies

lelanthranyesterday at 7:19 PM

> I kind of think there's room for a new dynamically-typed language that is designed around being fast to execute and doesn't cost such a huge performance multiple right off the top, and starts from day 1 to be multi-thread capable, but on the whole the trend is clearly in the direction of static typing.

Other than the "new" qualifier, Lisp supports all of that - SBCL compiles to native code, ecl/gcl compile to C (IIRC), etc.