Re: Switch/Case statements revisited

["steve donovan" <steve.j.donovan@...>]

On Nov 9, 2007 3:21 AM, Norman Ramsey <nr@eecs.harvard.edu> wrote:
> I'm not sure of the current state of token filters; I will be
> interested to see if anyone succeeds along those lines.  (Although I
> must confess that token filters remind me unpleasantly of the failed
> experiments with extensible syntax in the 1960s.  I would much rather
> have a single case syntax that is blessed by the Lua team.)

Absolutely. There's the perenial danger of 'private languages'. This
is always a potential problem in languages which are very expressive,
like C++. There the consensus is that the class designer is assumed to
be an intelligent and mature adult (;)). Style and idiom become very
important!

The main point of token filter experiments is to implement (a)
features we would like to prototype without messing around with the
core (b) domain-specific extensions. Lua is a very good language for
writing DSLs (or 'little languages' as the old-style Unix people
called them).

There's another power of specialized idioms; you can express the
problem more concisely so you can fit the program in your head:

http://www.paulgraham.com/head.html

steve d.