|Date||Sunday the 29th of September at 13:00 CET|
|Participants||PGL (chair), NB, KP, NP, JWC; Minutes by NB.|
The action item list is maintained as a tracker on SourceForge.
It was noted that the LB minutes incorrectly state that there are no open RMs, when there is one remaining in Discussion (RM#19). NB has made some progress with the VDMJ implementation of RMs #18 and #20. PGL and KP volunteered to look at some of the issues arising from the implementation, while MV is very busy.
From SS’s email, there has been no change in VDMTools. It was noted that Mr. Tomohiro Oda created a project for VDMPad on SourceForge.
The following changes were made to the 1.x.x series branch. I think most of these have been ported to the 2.x.x branch for release now:
2013-09-13 Correction to traces running in VDM-RT
2013-09-05 Clean up threads properly with exit statements
2013-08-14 Correction to implicit function satisfiability POs
2013-08-14 Correction to allow self to be used in thread sections
2013-08-12 Efficiency improvement for traces that create objects
2013-08-09 Small correction to empty sequence subtype testing
2013-08-09 Fix for polymorphic parameter type comparisons
2013-08-09 Fix to identify duplicate trace definitions in SL
2013-08-04 Correction to pattern matching in set binds
2013-08-01 Improvement to state designator error messages
2013-08-01 Fix for breakpoint on exit statements with no argument
2013-07-31 Type resolution correction for value definitions
2013-07-26 Correction for old~ names in postcondition apply expressions
2013-07-03 Correction to subtype POs for implicit functions
2013-07-01 Improve ParameterPatternObligation and add missing param pattern checks
No update here. PGL will discuss plans with Augusto.
Code-wise, the release is ready to go. What remains is all of the ‘bookkeeping’ effort of ensuring that everything is bundled correctly, and so on. JWC’s time in September was occupied by teaching and COMPASS deliverables, so the release still needs to be done. With a little bit of luck, it will happen next week. Version 2.0.0beta6 (on SourceForge) is the version we will release as 2.0.0. This version is Java 6 compatible.
At some point in October we expect to release a v2.0.2 (odd final numbers → development, even final numbers → release), that includes some cleanups small fixes.
Remaining work includes:
Version 1.2.5 is now up on SourceForge. As noted in an email to overturetool-core, JWC hopes that this is the last release in the 1.2.x series. The build structure of 1.2.x is far behind that of the 2.x series, and consequently is harder to produce a new version in it.
AstCreator split out
AstCreator has been split out of the main Overture repository and is now hosted in a repository on GitHub, and there is a corresponding maven repository there to access the libraries. This allows us a bit more flexibility in the build, and means we do not need to keep Overture 100% in lockstep with changes made to AstCreator. It’s also giving us a nice test of how suitable GitHub is. We will be publishing AstCreator releases to the main Sonatype maven repository, so that no special configuration would be required for anyone to use AstCreator in their own projects.
We have a new buildserver at AU for the Overture (and derivate projects). It’s still in the process of being configured; it probably needs 1–2 days of effort to move everything over. The switchover should be invisible to most people, however.
The Strategic Research Agenda is reviewed every other NetMeeting.
There were considered too few people at the NM to discuss this, especially without JF and MV. PGL will start a thread discussion by email to seek contributions from the group as to the way forward.
See Planned Publications. This has been updated by PGL, but no other changes were identified at the meeting.
PGL’s new Overture video was discussed briefly and was appreciated by all. Some suggestions were made for using full-screen display in places, and the possibility of showing the UML conversion tool was discussed. There was a brief discussion about whether it is better to show “incomplete” features and admit their shortcomings, or just to not mention what the tool cannot yet do.
We noted that no progress has been made on plans for the next Overture workshop. There was not enough support for a workshop in Singapore at FM2014. We need to make sure this is not forgotten about.