In C, you can use macros and functions in such a
way that enhances your productivity in implementing J.
The APL90 paper described the implementation of ,:
as follows:
Fl(laminl){R reshape(over(one,shape(w)),ravel(w));}
F2(lamin2){R over(a,reshape(over(one,shape(AR(w)?w:a))
,ravel(w)));}
I don't need the C++ features. My impression (could be
wrong) is that C++ compilers are more widely available.
----- Original Message -----
From: Dan Bron <[email protected]>
Date: Wednesday, July 22, 2009 10:08
Subject: Re: [Jprogramming] Reimplementing J
To: Programming forum <[email protected]>
> Roger wrote:
> > First choice: J; second choice: C.
>
> Ambrus concurred:
> > I'd actually use C, not a high level language.
>
> Why C?
>
> I only ask because, in my mind, C involves a lot of low-level
> housekeeping. To me, C's advantage is performance (and, to
> an extent, portability),
> not productivity. Unless you're thinking of specific tools
> or frameworks
> that would mitigate the housekeeping a bit?
>
> > For the second, it'd be ok to use a C++ compiler
> > but I would restrict myself to only the C features.
>
> Why C and not C++? And why would I use a C++ compiler if I
> was coding pure
> C?
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm