[plt-scheme] Why do folks implement *dynamically* typed languages?

On Wed, May 30, 2007 at 11:37:17AM -0700, Michael Vanier wrote:
> I find this analogy quite amusing, since every time I use PowerPoint I have 
> to manually turn off all the autocorrect options because it invariably 
> corrects things in ways I don't want.  I find myself saying "just do what I 
> want!" over and over again.  Apparently other M$ products do the same; I 
> had a friend write me an email about improv comedy where all the "improv"s 
> were changed to "improve" -- after all, everyone knows that "improv" isn't 
> a word, right?
> 
> BTW one aspect of this discussion that Matthias alluded to but didn't 
> describe in detail is languages with dependent type systems e.g. Epigram.  
> In those languages, the type system is extremely powerful but also 
> undecidable, so the programmer takes on the proof obligations him/herself.  
> My impression is that advanced statically-typed languages are moving in 
> this direction, which may make the boundary between such languages and 
> Scheme very small indeed.

I was working on this stuff in the 80's.  It's neat stuff.

-- hendrik