Potential problem with AC-Vector-Fancy package
Is anyone maintaining the AC-Vector-Fancy package? I haven't had a reply from the latest maintainer (Andrew Coppin) on Hackage, so I thought I'd open it up to cafe: I think I have found a problem with the union function: If you look here: http://hpaste.org/49889 You will see that line 4 gives a different result to lines 6, 8, 10; this shouldn't be the case because union is commutative.
Is anyone maintaining the AC-Vector-Fancy package? I haven't had a reply from the latest maintainer (Andrew Coppin) on Hackage, so I thought I'd open it up to cafe: I think I have found a problem with the union function: If you look here: http://hpaste.org/49889 You will see that line 4 gives a different result to lines 6, 8, 10; this shouldn't be the case because union is commutative.
On 10/08/2011 11:04 PM, Dave Tapley wrote:
Is anyone maintaining the AC-Vector-Fancy package? I haven't had a reply from the latest maintainer (Andrew Coppin) on Hackage, so I thought I'd open it up to cafe:
Oh, right. I haven't checked my mailbox recently...
I think I have found a problem with the union function: If you look here: http://hpaste.org/49889 You will see that line 4 gives a different result to lines 6, 8, 10; this shouldn't be the case because union is commutative.
AC-Vector-Fancy is merely a "fancy" facard over AC-Vector. So the bug is actually with AC-Vector. Looking at my source code, the true bug is in Data.BoundingBox.Range [which provides the engine that all the other bounding box types use). The actual bug turns out to by face-slappingly stupid: it's a typo in one of the variable names. I'll go get that fixed... and then maybe write some QuickCheck properties.
I think I have found a problem with the union function: If you look here: http://hpaste.org/49889 You will see that line 4 gives a different result to lines 6, 8, 10; this shouldn't be the case because union is commutative.
AC-Vector-Fancy is merely a "fancy" facard over AC-Vector. So the bug is actually with AC-Vector.
Looking at my source code, the true bug is in Data.BoundingBox.Range [which provides the engine that all the other bounding box types use). The actual bug turns out to by face-slappingly stupid: it's a typo in one of the variable names.
I'll go get that fixed... and then maybe write some QuickCheck properties.
I just updated AC-Vector 2.3.2, which fixes the bug. Sorry about that... Let me know if you find any other stupid mistakes. (Or even clever ones, but I rather doubt that!)
participants (3)
-
Andrew Coppin -
Dave Tapley -
Dave Tapley