At 01:33 PM 8/9/00 -0400, Chaim Frenkel wrote:
>I don't think it will be a half million objects.

The problem I have with this RFC is that it is too broad a brush.  Since 
different core functions would need to return different objects (there 
would be some common ones, but otherwise it would be pretty near 1:1, I 
would think), what we're really saying here is that we're overloading the 
interface for each core function differently.

Has the author really looked at every core function and determined an 
object for each?  How about exists()?  defined()?  sin()?  chr()?  There 
are surely some functions for which it doesn't make sense to come up with 
an object interface just for the sake of it, so that word 'all' bugs me.

I'm not against it in principle for many functions.  In fact, it could be 
used to improve RFC 52 by overloading, e.g., chown:

     $result = chown $uid, @dirs;
     print "Failed for these dirs: ", $result->failures;
     print "Failure reasons: ", map $_->failure_reason, $result->failures;

But IMHO that word 'all' puts a stake in its heart.  Each system function 
should be evaluated on its own merits.

--
Peter Scott
Pacific Systems Design Technologies

Reply via email to