Hi,
I found the problem and have reported it to the Freehep team. I could actually fix it myself in the Freehep SVN rather easily but I'm checking in with them first in case there are unintended side effects.
The issue is that the method createPlotterFactory() which has a name argument does not properly override the superclass method on the default AnalysisFactory implementation. This means that a DummyPlotterFactory is returned from this method instead of the proper implementation, and hence the DummyPlotter and DummyPlotterRegion objects are created instead of the real ones.
This explains the problem reported by Sho where DummyPlotterRegion objects are sometimes showing up in analysis code. When a "named" PlotterFactory is requested, e.g. the way it is done in most of the monitoring Driver code, this bug will show up. It explains why the IPlotterStyle objects are null on the regions, because the DummyPlotterRegion does not create them.
The JfreeChart backend doesn't have this bug because its implementation of AnalysisFactory properly overrides both methods.
I'll let you all know when this is fixed.
--Jeremy
-----Original Message-----
From: McCormick, Jeremy I.
Sent: Wednesday, April 02, 2014 11:44 AM
To: Johnson, Tony
Cc: Turri, Max; Uemura, Sho
Subject: AIDA issue
Hi,
I'm not sure this a bug but I found some interesting behavior in AIDA.
When you call the versions of createPlotterFactory() and IplotterFactory.create() which take a name argument, they return the "Dummy" versions of the classes, e.g. DummyPlotterFactory and DummyPlotter.
When the versions of these methods are called that have no arguments, the proper implementation classes are returned.
This doesn't seem correct to me. Shouldn't both versions of these methods return the non-dummy classes?
--Jeremy
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1
|
