develooper Front page | perl.perl5.porters | Postings from August 2010

Re: Patch to make string-append on win32 100 times faster

Thread Previous | Thread Next
From:
demerphq
Date:
August 24, 2010 10:39
Subject:
Re: Patch to make string-append on win32 100 times faster
Message ID:
AANLkTins6EuwiRTyfJ=-BNJ=8TKiKaMhvoNqk=AXXfpC@mail.gmail.com
On 24 August 2010 19:19, Wolfram Humann <w.c.humann@arcor.de> wrote:
> On 21.08.2010 13:46, demerphq wrote:
>>
>> On 21 August 2010 12:52, Wolfram Humann<w.c.humann@arcor.de>  wrote:
>>
>>>
>>> perl -MData::Dumper -E"sub mt{my $d=shift;$d>0?[mt($d-1),mt($d-1)]:int
>>> rand
>>> 100}; say length Dumper mt 18"
>>> 34576966
>>>
>>> needs approx. 6 seconds to print and 8 seconds to finish on both of
>>> these.
>>>
>>
>> Interesting, how does that compare for you if you stick a
>>
>>   $Data::Dumper::Useqq=1;
>>
>> before the dump?
>
> Some more results, wristwatch timed -> ignore minor differences:
>
> N=20
> strawberry:    27 seconds
> with my patch: 29 seconds
> usemymalloc:   10 seconds
>
> With $Data::Dumper::Useqq=1 I had to reduce N to 19 to avoid out-of-memory.
> Strawberryperl unpatched only.
>
> Without $Data::Dumper::Useqq=1: 10 seconds,  462 MB
> With $Data::Dumper::Useqq=1: 28 seconds, 1100 MB

Interesting...

>>> No noticeable difference. With a touch of horror I have to report that
>>> the
>>> printed stringlengths are slightly non-deterministic from run to
>>> run...???
>>>
>>
>> :-)
>>
>> No worries there, the code does use rand() after all, so we expect to
>> see slightly different results from run to run (or the srand
>> initialization would be broken). Add a srand(1) or change the int rand
>> 100 to a constant and it should produce repeatable results.
>>
>
> O.k. so no need to be worried about the code. But maybe I need to be worried
> about a mind that is obvously able to type "rand" one moment and complain
> about the code running non-deterministic about 5 minutes later ;-)

I suspect your holiday will resolve any concerns you might have in
this regard....

> I'll be on vacation soon. Won't be able to respond to news in this thread
> quickly (maybe not at all) for the weeks to come.

Have fun... And dont worry about this thread, seems to me the point is proved...

cheers
Yves

-- 
perl -Mre=debug -e "/just|another|perl|hacker/"

Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About