> That's not how I would "fix" this, if I were to. The transform's
> constraint should be (constant-arg (satisfies proper-list-p)) and
> proper-list-p should be made foldable.
Yeah, I was trying to avoid proliferating the number of calls to
proper-list-p (there's already one in memq-translation-as-case), but
probably that's not worth bothering about.
> However, to me this is a won't-fix and I feel pretty strongly about
> that.
Strongly enough to veto a change? Strongly enough that, if it had been
written as in my patch in the first place, you would take out the checks
and introduce the error at compile-time?
Douglas Katzman <email address hidden> writes:
> That's not how I would "fix" this, if I were to. The transform's
> constraint should be (constant-arg (satisfies proper-list-p)) and
> proper-list-p should be made foldable.
Yeah, I was trying to avoid proliferating the number of calls to n-as-case) , but
proper-list-p (there's already one in memq-translatio
probably that's not worth bothering about.
> However, to me this is a won't-fix and I feel pretty strongly about
> that.
Strongly enough to veto a change? Strongly enough that, if it had been
written as in my patch in the first place, you would take out the checks
and introduce the error at compile-time?