[Pharo-project] Additional Test for Shared Queue (s)
While trying to track down a bug, I started thinking that there was a problem with SharedQueue. There isn't, but there were no tests that convinced me of that. So I wrote one that actually has two processes reading and writing many times to a Q. The same tests works for NSSharedQueue (except that the class under test is hard coded). I suggest that putting these tests in the appropriate packages/core would be a good idea. Tests attached.
Thank you Andrew for your report. I open an issue for this: http://code.google.com/p/pharo/issues/detail?id=4181 On Sun, May 8, 2011 at 5:59 AM, <pharo-project-owner@lists.gforge.inria.fr> wrote:
While trying to track down a bug, I started thinking that there was a problem with SharedQueue. Â There isn't, but there were no tests that convinced me of that. Â So I wrote one that actually has two processes reading and writing many times to a Q.
The same tests works for NSSharedQueue (except that the class under test is hard coded). Â I suggest that putting these tests in the appropriate packages/core would be a good idea.
Tests attached.
-- Serge Stinckwich UMI UMMISCO 209 (IRD/UPMC), Hanoi, Vietnam Every DSL ends up being Smalltalk http://doesnotunderstand.org/
On 08 May 2011, at 00:59, pharo-project-owner@lists.gforge.inria.fr wrote:
While trying to track down a bug, I started thinking that there was a problem with SharedQueue. There isn't, but there were no tests that convinced me of that. So I wrote one that actually has two processes reading and writing many times to a Q.
The same tests works for NSSharedQueue (except that the class under test is hard coded). I suggest that putting these tests in the appropriate packages/core would be a good idea.
Tests attached.
Excellent approach, Andrew. Sven
tx! I want more of such tests. :) On May 8, 2011, at 12:59 AM, pharo-project-owner@lists.gforge.inria.fr wrote:
While trying to track down a bug, I started thinking that there was a problem with SharedQueue. There isn't, but there were no tests that convinced me of that. So I wrote one that actually has two processes reading and writing many times to a Q.
The same tests works for NSSharedQueue (except that the class under test is hard coded). I suggest that putting these tests in the appropriate packages/core would be a good idea.
Tests attached.
<NSSharedQueueTest-testCounting.st><SharedQueue2Test-testCounting.st>
participants (4)
-
pharo-project-owner@lists.gforge.inria.fr -
Serge Stinckwich -
Stéphane Ducasse -
Sven Van Caekenberghe