![[Home]](/~alokem/funny-dummy-cropped.jpg)
Thesis/FeedbackThesis/Feedback|
I like the outline so far! You are being too modest! You can claim more contributions: * there is no other P2P framework that can be instantiated and costumized by the end user to address the needs of a specific file-sharing community. * depending how far you go with this, if you are indeed able to integrate with JXTA, you can claim that U-P2P adds a "document layer" on top of a "protocol layer", thus separating concerns. * in the case of a Gnutella deployment, creating communities optimizes the message broadcasts, since the neighbours depend on the community. (well to claim this we would probably need measurements. This could be too much work for your thesis). * your case study with Design Patterns addresses the well-known problem of sharing and discovering Design Patterns! (find refs for this): can expand into a whole chapter: state that it is a well-known and worthwhile problem, what are the current approaches, how U-P2P could help solve it, finding a Schema, limitations and successes of your approach, using this to validate U-P2P and detect weaknesses...... * Here's one more for you, which could help explain why our approach is typically suited to P2P systems (I'm reacting to the RESH review): you are proposing a P2P-based process for selecting a Schema for document description, since the selection/survival of a schema follows the same "darwinian" process as for any shared file. The "best" schema is the one that is being downloaded and shared the most. |