Marcus,
Don't roll it back.
I just like make one note.
If class implements interface with constructor then all its subclasses may
not change constructor's prototype.
If this is OK for all, then I haven't any other objections to the pacth.
Please invite me into IRC, if you will discussing something releated to ZE
and/or PHP language not in @internals. (I am available in MSN, but not
always able to look for IRC discussions).
Thanks. Dmitry.
> -----Original Message-----
> From: Marcus Boerger [mailto:[email protected]]
> Sent: Monday, March 06, 2006 10:54 AM
> To: Dmitry Stogov
> Cc: [email protected]
> Subject: Re: [ZEND-ENGINE-CVS] cvs: ZendEngine2 /
> zend_compile.c php-src/tests/classes
> ctor_in_interface_01.phpt ctor_in_interface_02.phpt
> ctor_in_interface_03.phpt ctor_in_interface_04.phpt
> interface_construct.phpt
>
>
> Hello Dmitry,
>
> Monday, March 6, 2006, 8:29:50 AM, you wrote:
>
> > Hi Marcus,
>
> > Is it your answer to my "break label" patch? :(
>
> Well that would be funny or childish even :-)
>
> > I didn't see final PDM's decision about constructor in
> interfaces. And
> > I didn't see any discussion about this, however may be I missed it.
>
> It was discussed once more on IRC and thus i simply put it
> here. Since it is a one line patch it can easily be dropped
> out again if there's a majority of people who are against this.
>
> > The question about constructors in interfaces is not
> simple, and both
> > points of view make sense. So I would like to see your and others
> > arguments?
>
> To me this makes perfect sense becuase there is no other way
> of enforcing a signature for class constructors. On one hand
> this is good because as we discussed a while back
> constructors should not have to inherit the signature.
> However having a way to reenforce this allows to write
> dynamic factories and stuff alike where a certain signature
> is important.
>
> Best regards,
> Marcus
>
>
>