In my opinion it would make sense to look also other formalism than feature models since their capabilities are restricted - focus on analysis level more than later stages. If we look industry cases perhaps among the most known would be Toshiba Software Factory (Matsumoto, Y.,, keynote at SPLC 2007) and Lucent telecom products (Weiss D., Lai., R, Addison-Wesley 1999) as the authors have written several appesrs on their work within these product lines. If you are looking for a wide variety of languages over very different product lines perhaps then you might be interested to look the paper we wrote few years a go to SPLC. We analyzed 20+ cases using domain-specific modeling languages that look also product derivation phase (available at http://users.jyu.fi/~jpt/TolKelSPLC2005.pdf).
I agree that feature models as trees are easier to visualiaze than other kind of models. I wonder are you interested to see various domain-specific, or better to say, family-specific languages that are richer and more expressive than feature models? Some of them are graphical and some focus more on tabular or matrix format.
Comment Wall (4 comments)
You need to be a member of Practical Product Lines to add comments!
Join Practical Product Lines
I think your topic is highly relevant and would like to know why you are focused on feature models as a visualization approach?
There are some industrial people here on the network, perhaps you can get them interested in your research.