I think my favorite experience would be when I
basically killed my own research project. One day, I noticed a particular
problem with our approach to merging a software architecture description
with a list of elements to add/remove at run-time. After some discussion,
we realized that our initial approach was not feasible and we had
to come up with a new approach. Although this did kill my research
project, it was also a very important discovery. I suppose this is
part of what research is about, discover what you can and cannot do.
I did, however, get moved to a new project and currently I am still
doing research with Dr. van der Hoek.