Dear Mike,
(This is my last post on this topic, so that Fearless Leader doesn't
come out speaking softly and carrying a big stick. :-)
> From: Mike T. Machenry
> Subject: Re: [Boston.pm] more syntax checking
> Date: Tue, Jul 19, 2005 at 11:35:53AM -0400
>
> This feature just plain doesn't exist is what everyone is saying. [...]
By "everyone" I guess you mean "Uri".
The fact that it doesn't exist is pretty clear. Jun Wan, Ben Tilly and
I would like to see such a stricture implemented, and we don't see any
theoretical or practical hurdles to implementing it.
Uri isn't saying that it doesn't exist. Uri is saying that it doesn't
make sense to have such a stricture because it won't work with dynamic
definitions and AUTOLOAD. My response is the same as before.
The stricture would definitely be incompatible with some legal Perl
code, but that is what strictures are all about.
> [...] I imagine that it's just harder to provide that option due to
> autoload and begin being in the language.
BEGIN blocks are not relevant to the discussion, IMHO. If there is
"eval q{sub junctive {print 'hello'}}" inside a BEGIN, it is just the
same as having "sub junctive {print 'hello'}" outside the BEGIN.
If "use strict sub_definitions" is incompatible with any module that
uses AUTOLOAD, that is a choice the user has to make.
> [...]
> if you are working in a setting where any static analysis at all is
> necessary perhapse Perl is not the language for you. [...]
I don't need static analysis. I just want perl to tell me at the end
of compilation whether there are sub invocations without definitions.
Anyway, I'll stop repeating myself right now to avoid excommunication. :-)
peace, || Udayachal: a newspaper edited by slum children:
--{kr.pA} || http://tinyurl.com/57jaf
--
This .sig intentionally left blank.
_______________________________________________
Boston-pm mailing list
[email protected]
http://mail.pm.org/mailman/listinfo/boston-pm
