ebXML   masters by research in Computer Science  by Sacha Schlegel
   http://www.schlegel.li/ebXML/index.html    is the homepage of this project.

 M E E T I N G

 
m e e t i n g . h e a d e r  
  
date:    March 31st, 2003
week:    14
report:    x out of y
location:    School of Computing at Curtin University
Building 314, Level 4 Meeting room Don Reid
person:    Don Reid, supervisor
Sacha Schlegel, student (2967)
 
m e e t i n g . a c t i v i t i e s  
  
  • Level 1 algorithm about to be finished (not 100% implemented, basics only implemented)
  • Matrix introduced to handling referencing conflict. Then matrix transformation for removing referencing conflicts.
    Possible optimizer not implemented.
            1) in x direction
               -  from
                 | a.1   | a.2 | a.3
            -----|-------|-----|---------
            b.1  |  2    |  1  |
            -----|-------|-----|---------
            b.2  |       |     |
            -----|-------|-----|---------
            b.3  |       |     |  4
            -----|-------|-----|---------
    
               - to
                 | a.1   | a.2 | a.3
            -----|-------|-----|---------
            b.1.1|  2    |     |
            -----|-------|-----|---------
            b.1.2|       |  1  |
            -----|-------|-----|---------
            b.2  |       |     |
            -----|-------|-----|---------
            b.3  |       |     |  4
            -----|-------|-----|---------
    
  • DeliveryChanels are other potential cases for the matrix problem solving.
  • Writing Transport elements to CPA. next is DocExchange. Then level one finished. Excluding Certificates.
  • Starting with level two framework this week. Framework includes: State Machine for negotation, messaging, diagrams.
 
m e e t i n g . n o t e s  
Messaging Responsibilty can be: a) user, b) program, c) underlaying transport protocol.
Business people interest: the more implemented, the more complete, the better.
Academic people interest: list key problem areas and deal with them (AccessAuthentication might be a nice problem case).
A strategy might be to monitor over many negotiations and then to try to figure out what people negotiate about and to try to find their preferences. Eg everybody uses Acknowledgements so we might rethink our implementation without Acks.
But first implement the negotiation framework/architecture, THEN think about strategies.
Build infrastructure with an interface first.
Then develop strategies which use that interface:
  • default strategy: pass everything to the user. Also useful to test underlaying infrastructure.
  • then potential strategy: implement one sample strategy that shows, that the framework is usable.
 
m e e t i n g . t o . d o . l i s t  
  
  
  
  
  
  


© Sacha Schlegel@2003!