|Date||05 November 2008, 1300 CET|
|Participants||Peter Gorm Larsen, JohnFitzgerald, Carlos Vilhena, Adriana Santos, Shin Sahara, Augusto Ribeiro, HKA Lintrup, Kenneth Lausdahl, Christian Thillemann and Nick Battle.|
The actions are all at Overture on SourceForge.
Overture Parser Update
Nothing to report.
Status of Adriana’s work on the Overture Test Automation Support
The current plan is to: Documenting the tool properly, and afterwards extend it for the constructs that are not yet being considered The future plan of my work is: Document how the tool can be used. There is a GUI which must be explained so that someone else besides myself can use it! The name of each trace should be changed for a path. In this way, the test cases that are automatically generated are saved in the path the user chose. Reduce the hand-made work that users must dedicate to use tool. Enlarge the type of test cases that can be automatically generated.
Status of Carlos’ work on the Overture Connector between VDM++ and JML status:
Stabilise the tool, in order to be able to work on its extension. Fix a number of bugs that need to be fixed before continuing the work. Carry through some extentions one of the more likely candidates will be VDM functions to JML abstract methods (and vice-versa) Add more information to the documentation, concerning the decisions I made
Status of Augusto’s work on the Overture Proof Obligation Generator
Proof obligations are now generated for polymorphic function but still needs extensive testing. Next step resolving the problem with mutual recursion between super-classes. The coverage test are up and running and polymorphic functions will be tested Next step is to study should study if the polymorphic functions can be transfered to HOL
New students (Kenneth and Hans Christian)
Looked into mapping the uml n-arry association to a instanca variable like: att1 : B * C * D where B,C,D are other classes (but only to enterprise architect). More information in the PDF-document email received
Status on Christian’s work
The eclipse plugin
The overall plan is to aim for a common interface for VDMTools and Overture functionality on the eclipse platform.
The MONDEX case study
no progress since the last NM, but look for more steps forward after the next Mondex day on 17 October. We are under a lot of pressure from Deploy High Command to spend effort on that. The latest MONDEX surge produced good improvement on the proofs, see table at: http://www.vdmportal.org/twiki/bin/view/Main/MondexCaseStudy Unit tests for the operations in the abstract world have been prepared but the tests for the operations in the concrete world are yet to be prepared. With the automated proof work, some progress has been made, but are still unable to replicate Sander’s results.
The Pacemaker case study
My presentation given on 15 September in London is now on the vdmportal Pacemaker page http://www.vdmportal.org/twiki/pub/Main/PacemakerCaseStudy/FitzgeraldPacemaker.pdf and going into the VSR repository. It addresses both the Z and VDM models. There was a very positive response, notable from Tony Hoare who was delighted to see a possibility of relating models in “alternative” formalisms. From the point of view of co-simulation, I have been in touch with some bio-computing people at Oxford who have managed to point me to some heart models. We will start with a single cell model and then advance to perhaps a systemic model but modern heart models are vast and based on difference matrices like a sort of finite element analysis approach. That area of co-simulation may be some way off. Who will do this? I can now say that the international team working on this area will include: us, Andrew Butterfield (Trinity College Dublin, Ireland), Dominque Méry (LORIA, France) plus a new PhD student, Steve King at York, of course Artur and Marcel in Brazil. I am planning a meeting of as many as possible of this group in Newcastle, now probably in the New Year. TrAmS?, one of the projects with which I am involved at Newcastle, is buying a one of the reference platforms for implementations of Pacemaker software.
No progress since last NM
Unfortunately neither Adriana, Carlos, Shin and John can make it.
The temporary plan of the workshop
José Nuno suggested Braga, in Portugal. Web-site is up here.
Recently Appeared (since last NM): None
There is increased interest in VDM Tools at Newcastle, particularly from the proof perspective. The name that we might hear more from is Alexei Iliasov. Alexei, who recently completed his PhD, is now working in TrAmS? on the definition of patterns for fault tolerance mechanisms. It turns out to be not very easy to incorporate such patterns into Event B (Alexei?s first formalism) and so we are looking at others, including VDM++ and CSP||B. I see VDM++ as an interesting possibility here, because the expression of genericity through o-o is something not available in the other formalisms. We of course lose out in the comparison because of the lack of proof support. I will arrange for Alexei to talk with both Peter and me during his forthcoming visit to Newcastle.
December 7th, 2008 - 13:00 CE