in reply to Re: AUTOLOAD and call tracing
in thread oop - obvious AUTOLOAD question

You are right in correcting my statement with your clogging
logs point. I was trying to express that debugging code does
not follow the normal rules of proper/good/correct coding.
It is equally easy to imagine a name-conflict where you
mistakenly call a defined routine thinking you were AUTOLOADing
it. Then instrumenting your code the other way could be
more helpful.
Or that your other thirty thousand AUTOLOAD calls result
in logging indirectly.

You gave me a laugh, thanks.

I don't believe you think that using caller to
implement access control or interface selection is a reasonable
idea. I don't believe that even though you just said it.

Replies are listed 'Best First'.
Re: Re: Re: AUTOLOAD and call tracing
by Sidhekin (Priest) on Sep 27, 2002 at 09:21 UTC

    I don't believe you think that using caller to implement access control or interface selection is a reasonable idea. I don't believe that even though you just said it.

    I know that kind of trick has been used to implement something like private or protected methods; I would rather not debate if it is reasonable or not.

    What I have done is use caller information to decide how a routine behaves, so that if caller()->isa("Picky::Class"), the data passed is handled ... slightly differently.

    What can I say? It seemed like a reasonable idea at the time.

    ... and it has worked without redesign for years now :-)

    The Sidhekin
    print "Just another Perl ${\(trickster and hacker)},"

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom