On 2020-12-02 12:54 +0800, Vladimir Nikishkin wrote:
>
> Wolfgang Corcoran-Mathe <xxxxxx@sigwinch.xyz> writes:
> >If it
> > provides a different integer seed on successive calls, isn't that
> > enough for this SRFI's purposes?
>
> It's not about seeding the random number generator. It is about
> measuring code performance, so it has to monotonically increase with
> time.

OK, thanks for clarifying.  I'm not sure why I assumed it was being
used as a seed generator.

> > (4) Is there a rationale for recommending that stream-null? and
> > the-empty-stream be defined as null? and (), respectively?
>
> Well... MIT/GNU-Scheme does it like this.

Unless there's a better reason than that, I suggest dropping those
recommendations.  Alternatively, the-empty-stream could just be
specified as an alias of (), and similarly for stream-null?.

> If I can bother you to test the `current=jiffy`-based one once again,
> please?

This test still fails under CHICKEN, and a similar expression also
fails on chibi:

    (check (> (- (runtime) (runtime)) 0) => #t)

The check here is inherently non-portable: only Schemes with slow
evaluation or large values of (jiffies-per-second) will give
different values for the two calls of runtime.  It should probably
be deleted.

--
Wolfgang Corcoran-Mathe  <xxxxxx@sigwinch.xyz>

"The composer makes plans, music laughs." --Morton Feldman