Re: PHP 4.0 Bug #5152: Object passed in a session generateserrors when member functions are called

Ok, now I don't understand at all. Wasn't Thies committing this fix just a couple a days ago, with this controversial commit? What was the argument all about if everything generally worked, pretty much? Zeev On Fri, 23 Jun 2000, Sascha Schumann wrote: > On Fri, 23 Jun 2000, Andi Gutmans wrote: > > > At 02:02 AM 6/23/00 +0200, Sascha Schumann wrote: > > >On Fri, 23 Jun 2000, Zeev Suraski wrote: > > > > > > > Right. The 4.0.1 behavior should match the 4.0.0 behavior; We should > > > > handle it right or not handle it at all... > > > > > > Andi: "We need to make the change in 4.0.1." > > > Zeev: "Right, 4.0.1 must behave like 4.0.0." > > > > > > You lost me. :) > > > > > > This is IMHO not such an important change that it would > > > warrant not releasing a stable and working 4.0.1. The new > > > code needs to be tested publicly. That means that I want to > > > commit it and give users a couple of weeks to test it. > > > > What Zeev meant is that if the change doesn't go into 4.0.1 we should roll > > back to the 4.0.0 behavior of it being an error. It will be too confusing > > for it to work in one mini-version of PHP. > > Hm. The change you refer to has happened in revision 1.26 and > 1.25 of var.c. With other words, this has been the case since > PHP 4.0b3. > > revision 1.26 > date: 1999/09/16 08:05:17; author: thies; state: Exp; lines: +37 -31 > php4 now deserialized php3 objects correctly! > php3 will ignore objects serialized by php4 (that's the price we have to pay!) > small fix in deserializing objects with static initializers (thanx boris) > ---------------------------- > revision 1.25 > date: 1999/09/15 15:03:06; author: thies; state: Exp; lines: +3 -2 > when unserializing classes: > if the class is not known the standard class (no methods) will be used. > if the class is known - the created object will functional again! > > - Sascha > -- Zeev Suraski <zeev@zend.com> http://www.zend.com/