[Mono-list] -- Math.cs --

2003-02-01 Thread yoros
Hi, I have sent a patch for Math.cs. Here I have a newer patch for this stuff. I returned to the original version of Math.cs, now the diff is more undestandable. The files are: mcs/class/corlib/System/Math.cs mono/mono/metadata/icall.c mono/mono/metadata/sysmath.h

[Mono-list] -- Math.cs --

2003-02-01 Thread yoros
Hi, I have a great, great problem with the last Math.cs I sent, forget it, please. Please get this one. Thank you, and sorry, again. Pedro -- Pedro Martinez JuliĆ” \ [EMAIL PROTECTED] )|[EMAIL PROTECTED] /http://yoros.cjb.net Socio HispaLinux #311 Usuario Linux #275438 - http:

[Mono-list] -- Math.cs --

2003-02-02 Thread yoros
Hi, Here are the final files that I got. I think that the Math performance is good and that the results are the most accurate that can be with "libm" because MS.NET are using another kind of numerical algorithms. I changed the following files: mcs/class/corlib/System/Math.cs mon

Re: [Mono-list] -- Math.cs --

2003-02-03 Thread dietmar
On Sun, 2003-02-02 at 09:42, [EMAIL PROTECTED] wrote: > Hi, > > Here are the final files that I got. I think that the Math performance > is good and that the results are the most accurate that can be with > "libm" because MS.NET are using another kind of numerical algorithms. > > I changed the fo

Re: [Mono-list] -- Math.cs --

2003-02-03 Thread Miguel de Icaza
Hello! > Do you have some test cases to test the behavior? We usually add simple > tests to mono/mono/tests. Such regression tests are really helpful and > makes sure that we stay compatible with the MS jit. I recently applied a patch that also affected Math.cs, and added my tests to the class li

Re: [Mono-list] -- Math.cs --

2003-02-05 Thread yoros
On Mon, Feb 03, 2003 at 11:35:48AM -0500, Miguel de Icaza wrote: > Hello! > > > Do you have some test cases to test the behavior? We usually add simple > > tests to mono/mono/tests. Such regression tests are really helpful and > > makes sure that we stay compatible with the MS jit. > > I recently

Re: [Mono-list] -- Math.cs --

2003-02-06 Thread dietmar
Marcus just pointed out the we cant use round(), because it does not conform to IEEE rounding conventions. Try the following test: using System; class Test { static void Main () { Console.WriteLine (Math.Round((double)2.5)); Console.WriteLine (Math.Round((

RE: [Mono-list] -- Math.cs --

2003-02-06 Thread Reggie Burnett
ary 06, 2003 4:54 AM > To: [EMAIL PROTECTED] > Cc: mono-list > Subject: Re: [Mono-list] -- Math.cs -- > > Marcus just pointed out the we cant use round(), because it does not > conform to IEEE rounding conventions. Try the following test: > > using System; > >

Re: [Mono-list] -- Math.cs --

2003-02-06 Thread yoros
In the documentation is said that Math.Round for doubles and decimals will do the bank rounding... It does now. See you, Pedro On Thu, Feb 06, 2003 at 11:53:56AM +0100, dietmar wrote: > Marcus just pointed out the we cant use round(), because it does not > conform to IEEE rounding conventi