[Insight-developers] Failing test

[Bill Lorensen]

This patch passes for me.

On Fri, Feb 17, 2012 at 3:51 AM, Richard Beare <richard.beare at gmail.com> wrote:
> Hi,
> I think I've corrected the KittlerIllingworth issue. I'm not certain
> about the other problems causing failing tests. The good news is that
> both failures on the ubuntu/debian machines result from a single grey
> level shift in estimated threshold. Most likely this is due to a
> slight difference in floating point precision. I've included a
> modification that may fix the issue, but I have no way of testing it
> at present.
>
> On Fri, Feb 17, 2012 at 9:21 AM, Richard Beare <richard.beare at gmail.com> wrote:
>> Hi,
>> I've been attempting to track down the problem, I have a guess to the
>> nature of one problem, but haven't been able to confirm, given the
>> lack of identical platform. The most recent changes to
>> KittlerIllingworth have been part of this effort. I will do the best I
>> can today (Friday, Australian time), but it may be that two of the
>> filters can't be part of the next rc.
>>
>> On Fri, Feb 17, 2012 at 8:14 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>> Richard,
>>>
>>> Do you have any suggestions for fixing the failing test for this patch:
>>> http://www.cdash.org/CDash/index.php?project=Insight&filtercount=2&field1=buildname/string&compare1=63&value1=ThresholdingErrorsFixes-Ibd02b86-4&field2=buildstarttime/date&compare2=83&value2=2012-2-16
>>>
>>> We want to do another release candidate tomorrow (Friday) but cannot
>>> do so with a test that fails on these three platforms. If you can't
>>> fix it, can you provide a patch that reverts to a working version?
>>>
>>> Bill
>>>
>>> --
>>> Unpaid intern in BillsBasement at noware dot com



-- 
Unpaid intern in BillsBasement at noware dot com