MiniD dsource project set up

Jarrett Billingsley kb3ctd2 at yahoo.com
Mon Jun 26 14:40:25 PDT 2006


"Rémy Mouëza" <ray.jay.ay.moueza at DoNtSpAm.gmail.com> wrote in message 
news:e7p9ui$1eoh$1 at digitaldaemon.com...

> Sorry, I don't want to be offensive, but I don't see lots of "interest" of 
> doing
> a scripting language so close to D, althought without adding any "new" 
> stuff
> that are typical of dynamic language.
>
> The way I see it : D has lot of strength and weaknesses. What about doing 
> a
> dynamic language that would make a kind of balance by supporting features 
> that
> would releave D where it kind of fails ? From what I have seen, this is 
> not what
> MiniD is planning to do. And maybe it is not what you intended to do.

You're absolutely right, and as I go along, MiniD might evolve into 
something far different from what it is now.

> First, I would either leave the strongly typing system for something more
> dynamic ( more like scripting language ), or keep it strong but would make 
> much
> more type inference, going beyond what D is doing, something looking like
> Nemerle.

A type-inferred language would be very interesting to write.  Though I'd 
kind of like to stick to some form of static typing, if nothing else as a 
proof-of-concept that a statically-typed scripting language can be 
higher-performance than dynamically typed ones.

> Second, I would implement some symbolic stuff or some kind of reflection 
> to
> manipulate the type system or make something like meta object protocol to 
> be
> able to programatically transform the MiniD programs. This could allow for
> aspect oriented stuffs and more.

Bizarre!

> Actually, I think that I am too demanding with a for-fun project. Maybe I 
> should
> keep quiet and implement my own Pythonish/Smalltalkish language in D.

Hehe :)  I'd like to see it.

> Anyway, D is a high level programming language, with features closed to 
> dynamic
> languages and high performance. I think it's the language of choice to 
> implement
> a scripting language. The more scripting language implemention we'll have, 
> the
> best we'll be able to prove it.
>
> MiniD is a good idea.
> Good luck !

You too! 





More information about the Digitalmars-d-announce mailing list