Re: Scripting language takes a silicon turn

[Enrico Colombini <erix@...>]

On Tuesday 24 January 2006 14:43, Boyko Bantchev wrote:
> First, it is not all the ``whitespace'' that is meaningful, 
> just indentation.

Unfortunately, 'indentation' is not well defined. You may encounter a source 
file (or section of file, or just a single line) using only tabs, another 
using only spaces, another yet using an impredictable mixture of both (some 
editors do the latter automatically). 
How do you measure 'indentation' when you are maintaining source written and 
changed by different programmers on different editors/OS using different 
settings and conventions at different times? What happens when you paste code 
in the middle of an indented block (having no clear block reference)?

Moreover, relying on whitespace (or, if you prefer, on indentation) makes it 
difficult to have code inside strings, which can be very useful in some 
scenarios. Mixing program-generated code from different sources could be even 
worse.

Last but not least, why have the programmer do manually something that can be 
automated by the editor? I'd rather have Pascal's BEGIN...END (which I 
cordially dislike) than no block delimiters at all.
If C has a design mistake, it's that it doesn't enforce the use of delimiters 
on conditional blocks. But I don't think removing all block delimiters is a 
solution (unless you happen to program on punched cards).

  Enrico

P.S. Sorry for taking the flame bait... this time I couldn't resist ;-)