Status: New Owner: ---- Labels: Type-Defect Priority-Medium Platform-Pharo Component-EyeSee
New issue 738 by alexandr...@gmail.com: bugs in EyeSee http://code.google.com/p/moose-technology/issues/detail?id=738
I played with the matrix in EyeSee in a Pharo image. I found some problems: - #new:withAll: and #new: seems to raise an emergency debugger, whatever the value we provide - Many references are made that do not belong to EyeSee. For example #zero that refers to MalMatrixCoercion
I also found other problems (e.g., hash). I added some tests for those. I produced version 0.9.2
Cheers, Alexandre
Comment #1 on issue 738 by alexandr...@gmail.com: bugs in EyeSee http://code.google.com/p/moose-technology/issues/detail?id=738
A plausible implementation of hash:
ESMalMatrix>>hash ^(self dimension hash bitXor: (self at: 1 and: 1) hash) bitXor: (self at: self dimension x and: self dimension y) hash
Updates: Status: Fixed
Comment #2 on issue 738 by tudor.gi...@gmail.com: bugs in EyeSee http://code.google.com/p/moose-technology/issues/detail?id=738
ESMalMatrix is an abstract class and the initialization calls a rowSize which is a subclassResponsibility. I do not know why the emergency evaluator happens, but it is not because of EyeSee.
In any case, ESMalMatrix should not be instantiated.
I also fixed the hash according to your suggestion.
Updates: Labels: Milestone-4.6
Comment #3 on issue 738 by tudor.gi...@gmail.com: bugs in EyeSee http://code.google.com/p/moose-technology/issues/detail?id=738
(No comment was entered for this change.)
-
moose-technology＠googlecode.com