Re: parent_ctor()

w>> I don╢t know why you (stas,andrei(not sure) and zeev (not sure too) ) w>> underrate PHP in a manner I don╢t understand. I think Kristian and the w>> others taking part at this thread won╢t force PHP to fit their personal w>> needs... I don't underrate it. I just say we don't need to put every feature in the world and the kitchen sink into the language. If some language has some feature (even a good fature), that doesn't mean yet that we should put it in. Or that it is feasible to put it in. There are a lot of features that are worth to add, but also a lot of features that can't be added without making PHP another language. w>> We should collect some info on the points of controversy and matching w>> right now. I╢ve the feeling that some points are discussed over and over w>> redundantly. So far I saw only 6 points: 1. Rename constructors to $SOME_DEDICATED_NAME - Need: don't see any; Feasibility: can be done; Backward compatible: no 2. Remove runtime-addable properties - Need: don't really see; Feasibility: can be done; Backward compatible: no 3. Remove methodless objects - Need: not any; Feasibility: cannot really be done without removing PHP functions; Backward compatible: no 4. Add keyword referring to parent - Done. 5. Private variables/method Need: don't see real need; Feasibility: can be done; Backward compatible: yes, if private access doesn't break execution, no otherwise. 6. Serialzie references - Need: claims to be useful; Feasibility: really not easy; Backward compatible: yes So, two last points are real candidates to implement, others most probably not. Do you have more point? Care to put them somewhere on webpage so we could refer to it? -- Stanislav Malyshev stas@zend.com +972-3-6139665