Re: automake and 'strict' Perl5

[Tom Tromey <tromey@cygnus.com>]

>>>>> "Rodolphe" == Rodolphe Ortalo <Rodolphe.Ortalo@cert.fr> writes:

Rodolphe> A patch does not really make sense. In fact, you need the whole
Rodolphe> file. I send a compressed one to you separately (54k).
Rodolphe> [BTW, Tom, do you want it?]

Not right now, since I can't use it without papers anyway.  When the
time comes, I'd prefer a patch based on the cvs version, even if it is
a very large patch.

Rodolphe> I don't know if these prototypes can have real bad
Rodolphe> side-effects (in fact, even though I'm aware of such
Rodolphe> potentially ill behavior, I've never observed it in
Rodolphe> real-life scripts). I tend to view them as a kind of
Rodolphe> comment, that happens to be understood by the perl
Rodolphe> interpreter - something that it sometimes confess to the
Rodolphe> user.

Is there a way we can get Perl to tell us when a function call might
change its meaning as the result of a prototype?  That way we could
see if they will be harmful for automake.

I suspect they may not be, since my Perl style, as far as I'm aware,
generally gives functions static argument types.

>> Additionally, Perl function prototypes are useless when writing OO
>> modules, which is more than likely the direction that Automake is
>> heading.

I think we are (eventually) headed for the OO stuff, so that's too
bad.  But if it would help catch bugs in the non-OO sections, it still
might be worthwhile.

Tom