in reply to Re^6: reset particular variable
in thread reset particular variable

I suspect that you are right. I will have to "noodle on this a bit" to completely convince myself, but I suspect that it is true.

As often happens in these threads, we have gone a long way from the original question about "reset"! I believe that there is consensus that package scoped variables and reset is not the best way although this is part of the language. I was surprised, but this IS truly a language construct. However, just because "it works", doesn't mean that it is a good idea - smaller scoped lexical variables is the "right" idea. Then we got onto map{} and how to suppress an "undef" value which is also a good idea.

Now we are onto something that hasn't come up in the last few years of my coding. Almost all my subs{} return something and that result is expected and used or I consider this a coding error. Subs that don't return things are signal handlers or subs that modify a reference that is passed in. Basically I don't try to check return values from subs that aren't expected to return anything as a lvalue and I don't have much experience checking for this situation in a normal sub.

I figure that until proved otherwise AnomalousMonk has it right! Great work! Thanks!

Replies are listed 'Best First'.
Re^8: reset particular variable
by Anonymous Monk on Aug 22, 2009 at 00:55 UTC
    I believe that there is consensus that package scoped variables and reset is not the best way although this is part of the language.

    reset is a holdover from perl4, no-one uses it anymore