Joerg Henne wrote: > > Stefan Koehler wrote: > > > > In general I'm with Giacomo here. I have no problem with > messages about creation > > and lookup of components. > > But I _have_ a problem with stacktraces of exceptions that > aren't realy exceptions > > (because they belong to the system's normal behavior) which are > really confusing. > > It makes your own debugging very hard to see the log flooded > with endless repeats > > of something like this: > > > > DEBUG (2001-11-28) 12:12.40:255 [cocoon ] > (/layout/edit-repository/asfd) Could not find component for role: > org.apache.cocoon.components.saxconnector.SAXConnector > > java.lang.NullPointerException > > this very special case is caused by a role definition without the > specification of a class which is present for SAXConnector. I > don't think that > this is a very good solution right now. > Furthermore, I understand that the AbstractEventPipeline checks for the > presence of a configured SAXConnector every time it is composed. > By adding a > simple flag for the negative lookup for the SAXConnector not only thos > spurious log messages but also a few CPU cycles could be saved. > When I implemented this test, the ComponentManager didn't support checking if a component is available or not. But in a later version the "hasComponent()" method was added to the ComponentManager, but unfortunately we forgot about changing our code. I fixed this ten minutes ago.
Carsten > Joerg Henne > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, email: [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
