Re: The semicolon habit (was: Q: Definition

Stephen J Bevan <bevan@cs.man.ac.uk> wrote:
>tim@handel.Princeton.EDU (Tim Hollebeek) writes:
> In that case, the same argument can be made _in favor_ of the
> semicolon habit :) If, for example, you edit with emacs and
> (setq c-auto-newline t), then you get automatic indenting and
> formatting, and still have the flexibility to indent the
> way you want when necessary. Languages that depend on
> indentation get to be a pain as you
> assymptotically approach the right margin. ...
>
>I've written a fair amount of Haskell in the past and never noticed
>the problem to which you refer. Do you have any examples from real
>programs where the identation has been a pain or is your argument
>purely theoretical?


No, actually, the language I was refering to was Python. I just
started using it extensively lately, and the
indentation-has-syntactic-value idea annoys me occasionally,
especially in deeply nested structures. Another complaint of mine is
that emacs can automatically indent my C code, but can't
automatically indent Python code because two programs that differ in
_indentation_ have different meanings. I find that editors that
indent automatically defeat the 'misleading indentation' problem
which is usually given as the main argument for indentation only, and
I find that the redundancy between indentation and the actual syntax
helps catch errors in such contexts.


I'm pretty sure it's entirely an issue of personal preference, tho :)
I just get annoyed when people suggest that indentation based syntax
is obviously superior.
---
Tim Hollebeek
--

Return to the comp.compilers page.
Search the comp.compilers archives again.