Re: [RFC] [VOTE] Filtered unserialize()

> On 3 Nov 2014, at 21:33, Stas Malyshev <smalyshev@sugarcrm.com> wrote: > >> Coming late to the discussion. Was there any discussion to make the >> new argument a callback instead? Pass it the fully-qualified class >> name, have it return true (the class should be loaded) or false (the >> class should not be loaded). Deprecate the >> unserialize_callback_func >> mechanism at the same time. > > That was not discussed. It can be made this way, though it would be more > complicated and have more moving parts, but that's not part of the > present RFC. The way the RFC does it doesn’t preclude using a callback, so this could always be done later. -- Andrea Faulds http://ajf.me/