Edi Weitz <[EMAIL PROTECTED]> writes:

> [EMAIL PROTECTED] (Gerd Moellmann) writes:
> 
> > On a second thought: If you're seeing type errors being detected,
> > and assuming you are running the tests with the same OPTIMIZE
> > settings, what you are seeing could be the type-check overhead.
> 
> What would they look like? I use the same optimization settings and I
> think I get the same compiler notes as before except for lots of
> messages looking like

I meant the type errors you got from something being NIL and not a
HASH, which you mentioned in another message.  The type-checks
themselves in PCL are done by wrapping stuff in (THE <slot-type> ...),
which should be sensitive to OPTIMIZE settings.

Reply via email to