[Mono-osx] bug in Random

[Duane Wandless]

Hello Mono-team... Can you please provide an answer?  Was the intention to
use a different algorithm for the Random class, or was that a mistake?
 Honestly I cannot see a technical reason to use a different one.
Thanks.

On Mon, Jan 19, 2009 at 5:43 PM, Rodrigo Kumpera <kumpera at gmail.com> wrote:

> Your application depends on an implementation detail of the MS runtime.
> There is no guarantee either that they won't change the Random algorithm.
>
> This is unfortunate to your application but the solution is pretty simple,
> copy mono's
> implementation to your project and rely on it instead of an external prng.
>
>
>
> On Mon, Jan 19, 2009 at 8:38 PM, MacUser <atlas at spiralgraphics.biz> wrote:
>
>>
>> This is also a point of concern for me.  My code , written on Microsoft
>> .NET,
>> relied on the fact that using the same seed will always produce the same
>> series of numbers.  To port to Mono, not only would I need Mono's Random()
>> to be consistant across all Mono platforms, but also consistant to the
>> results that Microsoft's Random() produces.
>>
>> By the way, I work in the algorithmic graphics generation area, where an
>> image produced with the same input values must always look the same
>> regardless of the platform the app is running on.  This is why reliable
>> seeding is so important.
>>
>>
>>
>> Rodrigo Kumpera wrote:
>> >
>> > It doesn't state the the algorithm has to be the same, just that it must
>> > be
>> > reproducible given the same seed.
>> >
>> > On Mon, Jan 19, 2009 at 3:27 PM, Duane Wandless <duane at wandless.net>
>> > wrote:
>> >
>> >> That is a valid question.  From MSDN, the Random class is a
>> Pseudo-random
>> >> number generator:
>> >> The random number generation starts from a seed value. If the same seed
>> >> is
>> >> used repeatedly, the same series of numbers is generated.
>> >>
>> >> http://msdn.microsoft.com/en-us/library/system.random.aspx
>> >>
>> >> The value in this is that given a seed will produce the same sequence
>> of
>> >> random numbers.  So if the client and server both know the seed then
>> >> the algorithm will work correctly on both platforms.
>> >>
>> >>
>> >> On Mon, Jan 19, 2009 at 12:09 PM, Matt Emson
>> >> <memsom at interalpha.co.uk>wrote:
>> >>
>> >>> Duane Wandless wrote:
>> >>>
>> >>>> Maybe I incorrectly assume that the pseudo-random number should be
>> the
>> >>>> same on both platforms... but I believe that it should be a
>> predictable
>> >>>> sequence given the same seed.
>> >>>>
>> >>> Maybe this is a silly question, but why would you want your *random*
>> >>> number to be the same on both platforms, more so "a predictable
>> >>> sequence"?
>> >>> Sure I'm missing something obvious, but... ?!
>> >>>
>> >>
>> >>
>> >> _______________________________________________
>> >> Mono-osx mailing list
>> >> Mono-osx at lists.ximian.com
>> >> http://lists.ximian.com/mailman/listinfo/mono-osx
>> >>
>> >>
>> >
>> > _______________________________________________
>> > Mono-osx mailing list
>> > Mono-osx at lists.ximian.com
>> > http://lists.ximian.com/mailman/listinfo/mono-osx
>> >
>> >
>>
>> --
>> View this message in context:
>> http://www.nabble.com/bug-in-Random-tp21544091p21553211.html
>> Sent from the Mono - OSX mailing list archive at Nabble.com.
>>
>> _______________________________________________
>> Mono-osx mailing list
>> Mono-osx at lists.ximian.com
>> http://lists.ximian.com/mailman/listinfo/mono-osx
>>
>
>
> _______________________________________________
> Mono-osx mailing list
> Mono-osx at lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-osx
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ximian.com/pipermail/mono-osx/attachments/20090125/c57a76d6/attachment.html