Blogs (9) >>
Sun 30 October - Fri 4 November 2016 Amsterdam, Netherlands
Sun 30 Oct 2016 11:00 - 11:30 at Matterhorn 3 - Run-time Reflection Chair(s): Stefan Marr

Nowadays, there are more and more applications that need to run uninterruptedly. This need requires minimizing the downtime to add new features or fix bugs. Dynamic Software Update (DSU) solutions allow updating applications while they are executing. A common concern in all DSU solutions is the migration of the application’s state. This migration should guarantee the coherence of the state between versions by either adding, removing, preserving or transforming state. In an object-oriented environment, this state is represented by instances. In this paper, we analyse the features that a DSU solution should have with the objective to understand the operations a DSU should provide. Our analysis focuses on the migration of instances. Then, we identify the Meta-Object Protocol (MOP) that a programming language should expose to support these operations. We scope our proposal to DSU solutions for class-based programming languages with eager and atomic instance migration. Finally, we validate our proposal with a prototype DSU implemented in the Pharo programming language using the identified operations.

Paper (meta16-final7.pdf)211KiB

Sun 30 Oct
Times are displayed in time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

10:30 - 12:10: Run-time ReflectionMETA at Matterhorn 3
Chair(s): Stefan MarrJohannes Kepler University Linz
10:30 - 11:00
A model for Reflection in Rule-Based Languages
Simon Van de WaterVUB, Thierry RenauxVrije Universiteit Brussel, Wolfgang De MeuterVrije Universiteit Brussel, Belgium
Media Attached File Attached
11:00 - 11:30
Instance Migration in Dynamic Software Update
Pablo TesoneInria Lille–Nord Europe, France Mines Douai, IA, Univ. Lille, France, Guillermo PolitoInria, Luc FabresseMines Douai, Noury BouraqadiMines Douai, Stéphane Ducasse INRIA Lille
Media Attached File Attached
11:30 - 12:00
AST Specialisation and Partial Evaluation for Easy High-Performance Metaprogramming
Chris SeatonOracle Labs
Pre-print Media Attached File Attached