On Mar 14, 6:09 pm, Marc Olschok <nob...@[EMAIL PROTECTED]
> wrote:
> John Doty <j...@[EMAIL PROTECTED]
> wrote:
> >[...]
> > Part of the problem here is Elizabeth's attempt to absolve code from
the
> > need to be readable on its own. Good comments are always a boon, but
> > it's better to have a situation where they add value to the code,
rather
> > than relying on them as a crutch to salvage unreadable code.
> Comments cannot do the latter anyway. The role of a comment is to
> explain _what_ is done by the particular piece of code --- not to
> tell _how_ the code does it.
Well, there are also comments that are notes to someone who may modify
the code later, like:
\ NB. .Niclwarn-line first because it resets Niclerr#
or, ``(1)'' in the code after stack noise, and:
\ (1) ( ca_after len_after len_original )
but they should be below the definition, not above.
|
