Karl, thanks for the explanation.
On Sun, Oct 10, 2010 at 1:58 PM, Karl Pauls (JIRA) <[email protected]> wrote: > > [ > https://issues.apache.org/jira/browse/CLEREZZA-320?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12919595#action_12919595] > > Karl Pauls commented on CLEREZZA-320: > ------------------------------------- > > @Hasan, there are certain cases where classes from the outside expect that > they can use any classloader to get to outside classes (e.g., swing). Thats > why we have a magic bootdelegation in felix that detects these cases and > bootdelegates automagically. This is one of the cases and I fixed it in the > issue for ipojo iirc. However, it might very well be that we don't detect > this special case because of your set-up. It looks like there is scala in > the mix too etc. so maybe thats what prevents us form auto-bootdelegation > (we only do it if we are sure - when in doubt we don't) > > > SSP rendering is slow: > scala.tool.nsc.interpreter.AbstractFileClassLoader.findClass needs a lot of > time to be executed > > > ----------------------------------------------------------------------------------------------------------------------- > > > > Key: CLEREZZA-320 > > URL: https://issues.apache.org/jira/browse/CLEREZZA-320 > > Project: Clerezza > > Issue Type: Bug > > Environment: in netbeans, during cpu profiling exclude all java > core classes > > Reporter: Tsuyoshi Ito > > Assignee: Reto Bachmann-Gmür > > Priority: Blocker > > Attachments: org.clerezza.test.render.zip, > page_reuest_wo_javacore.png > > > > > > during profiling session we detected that rendering ssp is very slow. A > lot of time is used to execute > scala.tool.nsc.interpreter.AbstractFileClassLoader.findClass. Seems that > each execution the classes are (re)loaded. IMO this should not be the case. > > -- > This message is automatically generated by JIRA. > - > You can reply to this email to add a comment to the issue online. > >
