Re: table.el

[Miles Bader]

Tak Ota <address@hidden> writes:
> It is perfectly legitimate the extender knowing exactly what it is
> extending and knowing how the original is implemented.

It may be necessary for a wrapper to have `inside knowledge' of the
function it is wrapping, but it is desirable to avoid it to the extent
to which it is possible to do so.  Since the exact names of parameters
have heretofore not been part of the defined interface of a function, we
shouldn't start to make them significant if it's not necessary.

Morever, I have a selfish reason to say this -- I'm currently
implementing lexical binding for elisp, and your mechanism _requires_
that a function use dynamic binding for its arguments.  If there were no
other way to implement it, then so be it -- those functions can't use
lexically-bound arguments -- but I don't believe it's necessary.

-Miles
-- 
Occam's razor split hairs so well, I bought the whole argument!