[Mono-bugs] [Bug 25046] Changed - Math.Log() not within Epsilon

[bugzilla-daemon@rocky.ximian.com]

Please do not reply to this email- if you want to comment on the bug, go to the
URL shown below and enter your comments there.

Changed by martin@gnome.org.

http://bugzilla.ximian.com/show_bug.cgi?id=25046

--- shadow/25046	Tue May 21 10:59:26 2002
+++ shadow/25046.tmp.31494	Fri Jun  7 17:22:28 2002
@@ -1,14 +1,14 @@
 Bug#: 25046
 Product: Mono/Class Libraries
 Version: unspecified
 OS: Red Hat 7.2
 OS Details: 
-Status: NEW   
+Status: ASSIGNED   
 Resolution: 
-Severity: 
+Severity: Unknown
 Priority: Wishlist
 Component: CORLIB
 AssignedTo: mono-bugs@ximian.com                            
 ReportedBy: ndrochak@gol.com               
 QAContact: mono-bugs@ximian.com
 TargetMilestone: ---
@@ -35,6 +35,26 @@
 		if (Math.Abs(a - b) <= double.Epsilon)
 			return 0;
 
 		return 1;
 	}
 }
+
+------- Additional Comments From martin@gnome.org  2002-06-07 17:22 -------
+Well, this test does not run "fine" on .NET - it's broken there
+since the "exact" value is (*)
+
+-0.832516953253036096072037253179587423801422119140625
+and not
+-0.83251695325303621.
+
+Looks like our corlib/runtime calculates the logarithm with a
+better precision than mscorlib - if you use the correct value
+-0.8325169532530361 for your b, then the test will pass with our
+corlib (but fail with mscorlib since it's getting rounding errors 
+there).
+
+So my suggestion is to set b = -0.8325169532530361 and check whether
+a-b <= 1e-14 - this'll work with both.
+
+(*) printf ("%50.48f", log (0.1234) / log (12.345)) in GNU Octave ...
+