On 02/10/2013 19:15, Bob DeRemer wrote: > > >> -----Original Message----- >> From: Mark Thomas [mailto:ma...@apache.org] >> Sent: Wednesday, October 02, 2013 1:28 PM >> To: Tomcat Developers List >> Subject: Re: 8.0.x / 7.0.x progress >> >> On 01/10/2013 19:39, Mark Thomas wrote: >>> 8.0.x trunk appears to be stable (i.e. no longer crashes) >> >> So far no-one has come forward with results to contradict this and folks that >> reported issues with RC3 have not been able to repeat them. >> >> The additional issues reported with the snake example have been resolved. >> >> Currently, the 8.0.x unit tests pass (with tomcat native built from >> 1.1.x) for BIO, NIO and APR/native. >> >>> but the >>> FormAuthenticator unit tests are very slow (24 minutes). This needs to >>> be resolved. >> >> This has been resolved. >> >>> I will port the latest APR changes to 7.0.x shortly. >> >> Also done. >> >>> A build of the latest 1.1.x tomcat-native code is required to achieve >>> that stability. The implication is that a 1.1.29 release will be >>> required before the next 7.0.x release. >> >> I'm now certain that a tomcat native 1.1.29 release will be required before >> we >> can roll the next 7.0.x release. >> >>> The unit tests have identified a rare failure mode in JULI during >>> context stop: >>> http://ci.apache.org/projects/tomcat/tomcat7/logs/1527504/TEST-org.apa >>> che.catalina.websocket.TestWebSocket.NIO.txt >> >> This does not appear to be connector related so it is very unlikely that any >> tomcat native changes would be required to fix this. >> >>> The fairly frequent (5%-10%) failure rate of TestCoyoteOutputStream >>> still needs to be addressed. >> >> This appears to impact all connectors so at this point it is unlikely a >> tomcat >> native change would be required. >> >> >> I think we are at a point where it makes sense to start a tomcat native >> 1.1.29 release. Mladen? > > Please excuse the possibly naïve question, but what does it mean to "start a > native release" - in terms of timeline? Is this a build/test cycle, or is > there more involved? Just curious what we may be looking at with regard to > impact on 7.0 GA.
It should just be a build, test and vote cycle followed by the same for 7.0.x. If an issue is found then it might take longer. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
