>
> On Mar 10, 2020 at 3:49 PM, <Benjamin Eberlei (mailto:kont...@beberlei.de)>
> wrote:
>
>
>
>
>
> On Tue, Mar 10, 2020 at 5:30 PM Mike Schinkel <m...@newclarity.net
> (mailto:m...@newclarity.net)> wrote:
>
> > > On Mar 10, 2020, at 11:28 AM, Benjamin Eberlei <kont...@beberlei.de
> > (mailto:kont...@beberlei.de)> wrote:
> > >
> > > Just to make sure you don't run in circles in this discussion thread
> > here, even when syntax is not fixed yet, it's not going to be a syntax
> > where the attributes are suffixed after the declaration. It would maybe
> > some other characters like %[Attr].
> > >
> >
> > It is your RFC so you are the arbiter.
> >
> > One final on syntax: Can I suggest you consider @:Attr?
> >
> > -Mike
> > P. S. Did you see the question on interfaces?
> >
>
>
>
> No I missed it. Attributes work on interfaces, but they are not inherited to
> parent classes. This is similar to how a class implementing an interface may
> have different docblocks. You can use the reflection api to get to interfaces
> from a class and check for their attributes there. For traits the attributes
> get copied into the using class.
>
>
>
>
I think you may have misunderstood my question. I was asking if using an
interface we could *require* a class to implement specific attributes. If yes
that would be extremely valuable.
-Mike
