On Wed, Jun 3, 2009 at 8:26 AM, Rajith Attapattu <[email protected]> wrote: > On Wed, Jun 3, 2009 at 5:44 AM, Gordon Sim <[email protected]> wrote: >> [email protected] wrote: >>> >>> Author: rajith >>> Date: Tue Jun 2 19:53:23 2009 >>> New Revision: 781141 >>> >>> URL: http://svn.apache.org/viewvc?rev=781141&view=rev >>> Log: >>> This is a fix for QPID-1887 >>> The TimeToLiveTest should be enough to cover this case. >> >> The TimeToLiveTest doesn't test the 0-10 codepath though, does it? If it was >> this error would have caused failures up until now would it not? > > That was also my thinking until I took a very close look yesterday. > Looking at the code I don't see a reason as to how it shouldn't have > failed. But when tried without the fix it also seems to work. However > this test has been failing before (according to Aidan). > I will look into this more closely. I want to ensure that this test is > acurate and covers this issue properly (All though on the surface the > test seems good).
That test is in the 010Exclude list.because, IIRC, it was failing every time with -Dprofile=cpp. I don't have qpid checked out on this machine (and I'm on holiday using rubbish coffee shop wifi so can't be bothered git cloning). It fails very intermittently against the Java broker, but that's not relevant here. - Aidan (who, btw, has arranged ASF beers in Cambridge for next Weds in Harvard Sq if anybody here's interested and not on [email protected]) -- Apache Qpid - Give me convenience or give me death http://qpid.apache.org --------------------------------------------------------------------- Apache Qpid - AMQP Messaging Implementation Project: http://qpid.apache.org Use/Interact: mailto:[email protected]
