Re: Bug in Kernel#method objects that call super?

Ryan Davis wrote:
> 
> On Jul 6, 2007, at 13:57 , Charles Oliver Nutter wrote:
> 
>> It seems incorrect for method objects to change the behavior of super. 
>> If I super in 'a', I want super's 'a' to be called, without exception.
>>
>> Can someone confirm this is a bug? In JRuby we always super up the 
>> same-named chain, so this represents an incompatibility.
> 
> This doesn't seem like a bug to me at all, and seems like it'll dredge 
> up a typing debate if anything.
> 
> In the areas where I've implemented 'super' behavior it was a dynamic 
> super. I can think of a number of scenarios where you'd desire that 
> (prototype based programming for example) and not the static super 
> behavior you're advocating. I'm guessing it is fairly arbitrary which 
> one we use, but if I got to vote, I'd say leave it dynamic.

Having super be dynamic breaks encapsulation in a bunch of ways. Of 
course Ruby doesn't really prevent you from breaking encapsulation, but 
this still seems really non-intuitive.

The biggest problem, in my opinion, is that in order for there to be a 
dynamic super you'd have to both alias *and* use method() on the alias. 
That's cumbersome, and getting a dynamic super out of that sure feels 
more like an ugly side effect than anything intentional. If it should be 
dynamic, it should have an easier mechanism and be explicit.

The other problem is that if a subclass's alias names a method that 
doesn't exist on the superclass, you simply *can't* use method() to 
invoke it. That's just plain broken. If you can call it directly, you 
should be able to method(x).call with the same behavior.

- Charlie