Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.programming.threads > #1225 > unrolled thread

About Parallel sort library....

Back to article view | Back to comp.programming.threads

Contents

  About Parallel sort library.... "aminer" <aminer@toto.com> - 2012-11-05 14:55 -0600
    Re: About Parallel sort library.... Gerald Breuer <Gerald.Breuer@googlemail.com> - 2012-11-06 22:25 +0100
      Re: About Parallel sort library.... Gerald Breuer <Gerald.Breuer@googlemail.com> - 2012-11-06 22:32 +0100
        Re: About Parallel sort library.... "aminer" <aminer@toto.com> - 2012-11-06 19:27 -0600
          Re: About Parallel sort library.... Gerald Breuer <Gerald.Breuer@googlemail.com> - 2012-11-07 03:21 +0100
          Re: About Parallel sort library.... Gerald Breuer <Gerald.Breuer@googlemail.com> - 2012-11-15 18:28 +0100
        Re: About Parallel sort library.... Noob <root@127.0.0.1> - 2012-11-16 11:46 +0100
      Re: About Parallel sort library.... "aminer" <aminer@toto.com> - 2012-11-06 19:25 -0600

#1225 — About Parallel sort library....

Hello,

I have stress tested Parallel Sort library but it didn't pass
the stress test under freepascal and lazarus , this problem
is caused by the freepascal compiler, not my library, but it did pass the 
test under Delphi.

I will advise you to use my parallel quicksort cause it has
passed the stress test under both freepascal and delphi
and i have modified parallel quicksort to avoid the worst
case performance and you can use it in an hybrid
way by passing ctmergesort to the constructor..


And you can download parallel quicksort and parallel sort library from:


http://pages.videotron.com/aminer/
?

Thank you,
Amine Moulay Ramdane 

[toc] | [next] | [standalone]

#1226

Am 05.11.2012 21:55, schrieb aminer (moron):

> Hello,
> I have stress tested Parallel Sort library ...

You're living in a different reality than those reading your posts
(or not). No one is interestet in your ridiculous trivial algorithm
libraries and even less people are interested in any refinements;
and even lesser because your code is written in a rather exotic
language.
Check your server-logs to see how many people download your code
to see that I'm right. And stop spamming this newsgroup with
announcements of tiny changes of your "library".

Look what real parallelization libraries like TBB look like. Then
you recognize what people dealing with parallelization really need.

[toc] | [prev] | [next] | [standalone]

#1227

Am 06.11.2012 22:25, schrieb Gerald Breuer:

> Look what real parallelization libraries like TBB look like.

Here's the TBB-documentation:
http://threadingbuildingblocks.org/documentation.php
Look for another job, aminer ...

[toc] | [prev] | [next] | [standalone]

#1230

You wrote:
> Here's the TBB

You can still do many things with threads and with threadpools,
that's what i did, and i am happy with that.



Amine Moulay Ramdane.





"Gerald Breuer" <Gerald.Breuer@googlemail.com> wrote in message 
news:k7bvlo$ebr$1@speranza.aioe.org...
> Am 06.11.2012 22:25, schrieb Gerald Breuer:
>
>> Look what real parallelization libraries like TBB look like.
>
> Here's the TBB-documentation:
> http://threadingbuildingblocks.org/documentation.php
> Look for another job, aminer ... 

[toc] | [prev] | [next] | [standalone]

#1231

First, you're looking to find attention. To get more, lower your
posting-frequency - it is just nerving.
And second I assume you're hoping that someone uses your work.
This can't be reached by releasing Pascal-code. Pascal is almost
dead. Learn C or C++ - or whatever. Ignoring that shows that my
opinion of you as being a moron is rahter appropriate.  And your
continous updates to code with might had been have bugs is sim-
plary inspiring confidence like verifying code with "stress tests".

[toc] | [prev] | [next] | [standalone]

#1233

Am 07.11.2012 02:27, schrieb aminer:

> You can still do many things with threads and with threadpools,
> that's what i did, and i am happy with that.

Threadpools make sense when the CPU-time consumed by the thread is
in the magnitude of the time to create a thread through the kernel;
when the CPU-time is a higher, creating a thread rather doesn't hurt.
As you can't know which amount of data your algorithms operate on
you can't know the runtime so you can't know if using thread-pools
is generally appropriate.

[toc] | [prev] | [next] | [standalone]

#1234

Gerald Breuer wrote:

> Here's the TBB-documentation:
> http://threadingbuildingblocks.org/documentation.php

ITYM http://threadingbuildingblocks.org/documentation

[toc] | [prev] | [next] | [standalone]

#1229

>Am 05.11.2012 21:55, schrieb aminer (moron):

Don't call me moron , a little bit of respect Sir.

>your code is written in a rather exotic  language.

It's a freeware that i wrote in Object Pascal, to help
the Object Pascal users , but if you don't like it, that's
your opinion , but don't call me moron.



Amine Moulay Ramdane. 

[toc] | [prev] | [standalone]

Back to top | Article view | comp.programming.threads

csiph-web