Ugo Cei wrote: > > Carsten Ziegeler wrote: > > Hmm, usually yes - but as only few people are debugging cocoon anyway, > > installing the checks keeps everyone happy and noone searches for the > > real bug. > > And there's also the problem that "assert" is available only in JDK 1.4, > unless we want to introduce some compatibility layer or such. So, my > question was more philosophical than practical ;-). > > Anyway, what are you going to do with my patch? May I suggest applying > it (this would at least fix the problems with redirects and the flow), > maybe logging a warning when those objects are null? And a comment, too. > Would you like me to produce a new patch with these mods? > Hmm, this is a difficult question - I currently don't have time to find the bug. But calling startProcessing/finishProcessing on the environment is not the correct solution. This should only be done when a new processing starts. So actually this is called in Cocoon.java and when a sub-request is started processing. So, the problem might be somewhere in the flow or in the redirector?
Carsten
