Our algorithm factors in the exception type and a normalized stacktrace (i.e. we attempt to strip the localized text and end up with just class, method and potentially file and line information). It does not factor in the message, since the message is likely localized and that would create a new exception for each language, even though it is actually the same exception.
This means that an exception of the same type with the same stacktrace is considered the same exception and so in your case a new exception with a different message would not trigger as a new exception.
We have loosely considered a way for you to affect this algorithm, for instance by providing a simple textual input alongside the exception that could factor in similar to the exception type, but we have nothing concrete on that yet.
Hope this helps in clarifying things
ICENIUM MONITOR AVAILABLE FOR DOWNLOAD.
Monitor and track applications built with Icenium at run-time. Reduce support by fixing run-time crashes as they occur, obtain user insights such as location and system configuration, find and remove application bottlenecks, and track which features are used most to help prioritize development.
Try Icenium with EQATEC Analytics today >>