On 9/23/07, Adam Harvey <aharvey php.net> wrote:
> Quoting Justin Patrin <papercrane gmail.com>:
> > This is exactly why I have always used =& in
my PHP4 code. You can
> > introduce subtle bugs if you don't use it, such as
if, in the
> > constructor, a reference to the object that is
being instantiated is
> > stored somewhere (such as in a sub-class). Various
PEAR classes *will*
> > break if code is changed to use = rather than
=&.
>
> I don't mind reversing the changeset (you know, if I
can figure out
> how to get CVS to _do_ that; that's reason #59394 why I
prefer
> Subversion right there) and pushing out 1.7.14 if
everyone thinks it's
> a bad idea, but give me some credit here: I did do an
unusually
> thorough round of testing on PHP4 after making that
change for
> precisely the reasons noted throughout this thread, and
as far as I
> can tell, it hasn't caused a problem for DB.
>
I wasn't saying that this had been a bad change, just
pointing out
that there had been a reason for all of us using =&
rather than = for
PHP4 code. If there are no regressions here then by all
means make the
change. Just be careful.
--
Justin Patrin
--
PEAR Development Mailing List (http://pear.php.net/)
To unsubscribe, visit: http://www.php.net/unsub
.php
|