This is the bit that gets me, I really want to see the spot in the mailing list archive or the IRC logs where they made this descision. The bit that really gets me with the references, is that they found a set of cases where PHP code caused the potential for a PHP crash, and their fix was to make the language more verbose by requiring temp userland temp variables, rather than fix the original problem internally. Why was that the better trade off?
Originally Posted by bonefry
Any idea what the dates were when they made these decisions? It would be a great help in understanding the situation to be able to piont at the archives and say "look there, that's why it happened". (If that is possible, it would be a great vindication of an open development process - you'd never exect to be able to see that sort of detail in a closed source project.)
Without that clarification, I don't see courage, I see internal bugs which the devs can't fix. And then snubbing outsiders when they question their decisions.
No, I take that back. After reading the internals list for a while, it does make sense. And the main problems come from writing PHP0.8 code, the best solution then is to not write PHP0.8 code. Bye bye PHP 4 support, I'm going to start migrating/trimming my code today, rather than trying to support 4 & 5 at the same time.