L
Leo Tohill
BitVector32 has a bug where it doesn't give the correct result for the 32'nd
bit. It always reports it as false. Disassembly of the method in
System.DLL reveals the cause of the problem - improper treatment of the sign
bit of the int32.
This was reported earlier, in a posting to
microsoft.public.dotnet.languages.vb on 2002-08-23 . However, no fix
appeared in 1.1. I'm surprised, given that this bug gives incorrect
results, not an exception, so there must be a few programs out there that
also also misbehaving due to this problem.
I'm posting now :
1) In case the previous report didn't get into Microsoft's tracking system.
2) To encourage a high priority for addressing the problem.
3) So that others encountering the problem may have an easier finding out
about it in a google search.
bit. It always reports it as false. Disassembly of the method in
System.DLL reveals the cause of the problem - improper treatment of the sign
bit of the int32.
This was reported earlier, in a posting to
microsoft.public.dotnet.languages.vb on 2002-08-23 . However, no fix
appeared in 1.1. I'm surprised, given that this bug gives incorrect
results, not an exception, so there must be a few programs out there that
also also misbehaving due to this problem.
I'm posting now :
1) In case the previous report didn't get into Microsoft's tracking system.
2) To encourage a high priority for addressing the problem.
3) So that others encountering the problem may have an easier finding out
about it in a google search.