Review queue

Why a queue and how to use it?

• The queue is intended as a TODO list for Malte. We hope that it can be a useful tool for Malte to keep track of who wants want and to keep the waiting times reasonably fair.
• Use the queue for code reviews, discussions and other Fast Downward things that Malte should work on.
• As of now, the queue is intended to be used by the Basel AI group only. If you think this should be changed, please get in touch.
• Add new items to the end of the queue. Usually, Malte will work from the front of the queue, but there may be reasons to promote/demote things.

• There is no need to put everything into the queue. It is intended for things that you would ideally like to get done right now, but that are not terribly urgent.

• The queue should be easy to use and useful for everyone involved. If it turns out that this is not the case, we should get rid of it again.

The queue

• http://issues.fast-downward.org/issue26: integrate preprocessor in search code (Gabi)

• http://issues.fast-downward.org/issue629: task interface for pruning techniques (code review, Florian)

• http://issues.fast-downward.org/issue509: task interface next steps (discuss long-term plan, Florian)

• http://issues.fast-downward.org/issue667: Integrate SCCs based merge strategy (code review, Silvan)

• http://issues.fast-downward.org/issue668: Integrate dynamic MIASM merge strategy (code review, Silvan)

• http://issues.fast-downward.org/issue688: replace GlobalOperator* by operator ids in successor generator and for preferred operators (code review, Florian, Jendrik)

• http://issues.fast-downward.org/issue693: new hash function (use mini-benchmark to reduce the overhead of new hash function, Jendrik)

• http://issues.fast-downward.org/issue719: clean up option parser (code review, Jendrik)

• http://issues.fast-downward.org/issue638: bugfix in systematic pattern generation (code review, Florian)

• http://issues.fast-downward.org/issue707: M&S: make pruning of unreachable and irrelevant states configurable (code review, Silvan)

• http://issues.fast-downward.org/issue693: new hash function (come up with implementation strategy, Jendrik)

• http://issues.fast-downward.org/issue705: Reduce memory usage of successor generator (code review, Florian)

• http://issues.fast-downward.org/issue725: Use OperatorID instead of GlobalOperator* more often (code review, Florian)

• http://issues.fast-downward.org/issue726: Change StateRegistry::get_successor_state from GlobalOperator to OperatorProxy (code review, Florian)

• http://issues.fast-downward.org/issue198: issues when using multiple heuristics (added by Malte, to be discussed with Salomé)

• http://issues.fast-downward.org/issue718: merge heuristic into evaluator (added by Malte, code review for Manuel)

• http://issues.fast-downward.org/issue717: Lama-Synergy: get rid of the synergy object (added by Malte, code review for Cedric)

• http://issues.fast-downward.org/issue694: new hash table implementation (code review, Jendrik)

• http://issues.fast-downward.org/issue668: Integrate dynamic MIASM merge strategy (code review, Silvan)

• http://issues.fast-downward.org/issue700: RootTask class (code review, Florian)

• http://issues.fast-downward.org/issue742: Fix a bug in M&S (code review, Silvan)