Message5844

Author malte
Recipients andrew.coles, erez, florian, jendrik, malte, silvia
Date 2016-11-30.19:16:01
Content
One small addition: the total numbers in the last row of Jendrik's table assume
that the number of open list entries and the number of generated states are of
the same order of magnitude. For tasks with many duplicates, it is possible that
there are arbitrarily more open list entries, so that the open lists can become
the dominating factor for memory usage.

If we pursue the goal of minimizing memory usage, it would make sense to
introduce measurements to the planner that compare the maximum number of open
list entries with the number of generated states. If it turns out that open
lists are sometimes the bottleneck, we can rethink our current (quite lazy)
duplicate elimination scheme.
History
Date User Action Args
2016-11-30 19:16:01maltesetmessageid: <1480529761.64.0.968808784318.issue213@unibas.ch>
2016-11-30 19:16:01maltesetrecipients: + malte, erez, andrew.coles, silvia, jendrik, florian
2016-11-30 19:16:01maltelinkissue213 messages
2016-11-30 19:16:01maltecreate