On Thu, 03 May 2001 15:39:02 EDT, Benjamin Sugars wrote:
>On Thu, 3 May 2001, Gurusamy Sarathy wrote:
>> I don't like making the programmer responsible for avoiding coredumps
>> in the guts.
>
>Neither do I, but at some point it becomes unavoidable. Consider %SIG, for
>instance.

I hope you're not saying designing things badly is unavoidable because
bad design already exists elsewhere in the code.  :-)

>However, if the implementation of your scheme can burden that
>responsibility in this particular case, great.  Is this something you
>envision for any 5.x version of perl, or is too late for that?

My best guess is that it is surely in the realm of the possible for
Perl 5.  If it doesn't break binary compatibility, maybe even 5.6.x.


Sarathy
[EMAIL PROTECTED]

Reply via email to