🚀 go-pugleaf

In article <11122@alice.UUCP> pereira@alice.UUCP () writes:
>In article <708@ecrc.de> micha@acrab4.UUCP (Micha Meier) writes:
>> The Pereira benchmarks used
>>for the same purpose are not well suited because they  often  measure
>>something different than they say.
>
>I've just looked at the comments in the source for my benchmarks, and nowhere
>is there a statement of what the various programs are intended to measure,
>so I don't see what evidence you have for your statement.

I'm sorry I didn't express myself more clearly. What I had in mind is that
when you run Pereira benchmarks, you get an output of the form

tail_call_atom_atom took (1.45 - 0.533333) / 2000 = 0.000458333 seconds/iteration
binary_call_atom_atom took (2.65 - 0.533334) / 2000 = 0.00105833 seconds/iteration
cons_list took (2.83333 - 0.533334) / 2000 = 0.00115 seconds/iteration
etc.

so without looking at the code and the comments there, one would assume
that the program measures what the predicate names say, because for most
of the benchmarks the names look self-explanatory.
On the other hand, the comments in the source only
say e.g. '% 19. Make 1000 asserts of unit clauses',
but one assumes that this is what is being measured by this benchmark,
at least it is how I've understood it, however this one also
measures list unification and tail-recursive calls.

--Micha Meier

--
USA     micha%ecrc.de@pyramid.com	MAIL	Micha Meier
						ECRC, Arabellastr. 17
EUROPE  micha@ecrc.de				8000 Munich 81
						West Germany