See below...

----- Original Message -----
From: Mark Slagell <mslagell / iastate.edu>
To: ruby-talk ML <ruby-talk / netlab.co.jp>
Sent: Friday, October 13, 2000 5:38 AM
Subject: [ruby-talk:5503] Re: 2 ideas from Haskell


> hal9000 / hypermetrics.com wrote:
> >
> > Hi,
> >
> > In article <39E609EF.4A7672D6 / iastate.edu>,
> >   Mark Slagell <ms / iastate.edu> wrote:
> > > Do either of these interest anyone:
> > >
> > > 1. a "literate mode" that assumes all lines in a script are comments
> > > unless the first column is a special character (Haskell uses '>').
> > >
> >
> > Hmmm... not on my Top Ten list of features. I think =begin/=end are
> > basically enough...
>
> I have to disagree there: the =begin/=end scheme distinguishes comments
> from code easily from the interpreter's standpoint but not from the
> reader's (who has to look around for delimiters); it makes it a little
> easier to write comments but in the end makes it harder to read them.

OK, I see your point.

> Allowing the option to swap the code/comments default doesn't have that
> first-glance-ambiguity problem, and is a way to facilitate _very_
> literate programming, where there are often more comments than code.
> This may be of little concern to most of you (rubyists seem to abhor
> comments as much as perlists do! yeah, flame away at me for that :-) but
> it also allows some interesting possibilities such as being able to feed
> something essentially like natual-language documentation to the
> interpreter, peppered here and there with bits of real code.

I can see where this might be of some value... but I *think* I might like
some
other way of distinguishing besides the file extension. Maybe a command line
parameter? Or maybe something embedded at the top of the file? For scripts
using #!, of course, they become almost the same.

Hal