[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Sun, 13 May 2001 04:33:56 -0700
From: Neil Jerram <address@hidden>
Date: 13 May 2001 10:32:50 +0100
Thanks for this. There seems to be a strong consensus now that we
should include docstrings in the lambda body, and solve any memory
consequences if and when they become important.
Any thoughts on appropriate syntax for documenting values and
bindings? In other words the analog of, in Emacs Lisp:
(defvar fill-regexp "\\([hoopy]+\\)"
"Docstring goes here.")
Except with the additional possibility to say whether the docstring is
a property of the variable - `fill-regexp' - or the value -
"\\([hoopy]+\\)" - or perhaps specify both docstrings in the same
well, i hereby exercise my right to foolishly backpeddle from my
previous consensus-oriented stated opinion... i now think docstrings
being part of the lambda is not a good idea after all -- not actually
standardized, somewhat ambiguous, not amenable to localization (gettext,
AFAICT), as well as being gratuitously memory consumptive. having said
that, i also think that realistically, people will do this out of habit
anyway, so we might as well formualate some plan for supporting it...
i hope this waffling indicates that i'm not qualified for docstring
processing (he says, hoping to shirk guile maintenance for a while and
get back to Other Stuff ;-)...
btw, `use-modules' update and documentation is in. feedback requested!