Re: namespaces, goops, etc.

[Mikael Djurfeldt]

Michael Livshin <address@hidden> writes:

> > 1. define-class - class redefinition protocol
> 
> so you give up consistency for magic that you don't even need, really.

Well, I can tell you that I'm not entirely happy about the current
solution.  But I really tried to get the least amount of magic with
preserved convenience of use.

The magic involved is that the expansion of `define-class' uses
`defined?'.  Otherwise it expands to plain Scheme code.

> let's see when the class redefinition protocol is used.
> 
> * interactively, when you re-load some file/module or redefine a class 
>   from the prompt.
> 
>   in this case, the development environment should have enough
>   information to notice that a *re*-definition is taking place, and it
>   may even warn you about it or ask for confirmation.

Hmm... to me this seems like more magic than the current situation.
Can you explain exactly how this would work?

> * non-interactively, in a running program.
> 
>   if this were up to me, I would _not_ invoke the class redefinition
>   protocol magically in this case.

I don't get it.  How would the current `define-class' be used
non-interactively in a running program?

> > 2. define-method - may need to define the GF
> 
> so it can define it in the same local scope.  I don't see any problem
> here.

Again, this sounds like magic to me.  Can you explain how the
expansion of `define-method' should look like?

If you have a suggestion on how to improve these macros which is still
convenient to use, I'm very interested.

Mikael