some assembly wanted

1 view
Skip to first unread message

Leopold Toetsch

unread,
May 30, 2005, 12:02:53 PM5/30/05
to Perl 6 Internals
IIRC it was already mentioned here: http://shootout.alioth.debian.org/
- the Computer Language Shootout.

It would be quite interesting to have Parrot figures too and see
benchmark timings decrease steadily ;-)

I've someone wants to take over that part:
- configure => perl Configure.pl --optimize
- make && make install ought to be ok to /usr/local ...

So we'd just need the benchmark code translated to PIR. I've here
harmonic.pir and ack.pir (the latter sucks badly)

leo

Matt Diephouse

unread,
May 30, 2005, 4:17:30 PM5/30/05
to Leopold Toetsch, Perl 6 Internals
Leopold Toetsch <l...@toetsch.at> wrote:
> IIRC it was already mentioned here: http://shootout.alioth.debian.org/
> - the Computer Language Shootout.
>
> It would be quite interesting to have Parrot figures too and see
> benchmark timings decrease steadily ;-)

How is this different from timeparrot? <http://xrl.us/f9oy>



> So we'd just need the benchmark code translated to PIR. I've here
> harmonic.pir and ack.pir (the latter sucks badly)

Adding .pasm, .imc, and .pir scripts to examples/benchmarks/ will
cause them to be benchmarked and graphed as part of timeparrot every
day.

--
matt diephouse
http://matt.diephouse.com

Leopold Toetsch

unread,
May 31, 2005, 12:38:45 AM5/31/05
to ma...@diephouse.com, perl6-i...@perl.org
Matt Diephouse <mdd...@gmail.com> wrote:
> Leopold Toetsch <l...@toetsch.at> wrote:
>> IIRC it was already mentioned here: http://shootout.alioth.debian.org/
>> - the Computer Language Shootout.
>>
>> It would be quite interesting to have Parrot figures too and see
>> benchmark timings decrease steadily ;-)

> How is this different from timeparrot? <http://xrl.us/f9oy>

The difference is comparison to ~60 different compiled and interpreted
languages (click "Sandbox").

>> So we'd just need the benchmark code translated to PIR. I've here
>> harmonic.pir and ack.pir (the latter sucks badly)

> Adding .pasm, .imc, and .pir scripts to examples/benchmarks/ will
> cause them to be benchmarked and graphed as part of timeparrot every
> day.

Yes I know. We should do that additionally to get figures vs time to see
how Parrot evolves.

leo

Matt Diephouse

unread,
May 31, 2005, 10:39:52 AM5/31/05
to l...@toetsch.at, perl6-i...@perl.org
Leopold Toetsch <l...@toetsch.at> wrote:
> Matt Diephouse <mdd...@gmail.com> wrote:
> > Leopold Toetsch <l...@toetsch.at> wrote:
> >> IIRC it was already mentioned here: http://shootout.alioth.debian.org/
> >> - the Computer Language Shootout.
> >>
> >> It would be quite interesting to have Parrot figures too and see
> >> benchmark timings decrease steadily ;-)
>
> > How is this different from timeparrot? <http://xrl.us/f9oy>
>
> The difference is comparison to ~60 different compiled and interpreted
> languages (click "Sandbox").

Okay. I wasn't sure that you meant to actually add it to the Shootout.
Looking at the site again, I found the FAQ and understand what you are
saying.

Reply all
Reply to author
Forward
0 new messages