[nem-en] Re: Scala: what you say about it?

[Andrey Khropov]

Kamil Skalski wrote:

> >> - usability of Nemerle macros in multi-language, large projects - in
> >> my currently developed application I have a Nemerle library + some
> >> macros and GUI client in C#.
> > 
> > I hope it will be possible to design some GUI in WinForms designer (maybe
> > even XAML) and get Nemerle code soon (I'm working on it).
> 
> Yeah, when I created new Windows Forms project in VS integration and
> opened Designer... VS just vanished ;)

Well, I together with some CodeDom functions managed to create an infinite
recursion :). Now it's fixed although the whole thing doesn't yet fully work.

> Well, as you probably noticed our support for implementing features
> and fixing bugs is not very fast. We are surely not abandoning the
> project any time soon (maybe if there is an equivalent language with
> type inference, pattern matching and macros, but with support from big
> company/community, I don't know C# 5.0 ?).
> I think that with current feature set of Nemerle we are already in a
> reasonably stable position - language has most of the useful built in
> stuff already there and with macros the limits for extensibility are
> quite far from us. :-)
> I would estimate Nemerle as being ahead from current mainstream
> languages for about 5-6 years, at least in the coolest features.
> 
> I think that now there comes the time for stabilizing the language
> itself and focus on:
> - fixing bugs

Agree

> - improvements in compiler / macros API

Agree.

> - improvements in performance

Agree ( yes, I'm a performance junkie :-) )

> So, the conclusion is:
> - Nemerle is a very good and useful language at the moment and I don't
> mind using it as replacement for C#/Java for a long time

Yes, I actually came to it when I wanted to find something less verbose than
C++  that I use.

First I considered OCaml because it's quite efficient but inability to overload
operators and some other things pissed me off.
Then I came to Python and I liked it except classes. And I also prefer static
typing.
And than I spotted Nemerle and it found it to have Ocaml's FP, Python's
conciseness, metaprogramming that I liked in C++ and usual OOP stuff done right
as well.

> - as Michal mentioned, Nemerle did not "catch up" very well yet,
> probably because of the current trends in computer science, maybe
> because of some of our controversial design decisions (keeping close
> to ML languages and insisting on static typing), but as those
> decisions drived us to create this language and are good in our
> opinion it is not something to be ashame of

I think it's not about Computer Science it's just almost nodody knows about its
existence (maybe because you're not working in MS Research :) ).
When I tell somebody about Nemerle they usually say: 
"Hey, it seems to be a pretty good language. Haven't heard of it before".

And it took Ruby about 9 years to "catch up" until Rails killer app was created.

> - the future of Nemerle is defined by the community, we will have the
> language / platform, which we will develop all together

-- 
AKhropov