Re: An elisp question

[Oleksandr Gavenko]

Nikolaj Schumacher пишет:
> Incidentally, if anything, testing for t is more correct.  :) The doc for
> `y-or-n-p' clearly states, it will return t for "yes".  It could be
> extended to return 'cancel without changing the interface.  Using any
> non-nil value as "yes" is indeed an assumption.  (Not that it's likely
> to break, or that I haven't made that assumption myself...)

Thanks, yes for some functions there may by side effect with
non-nil as true!