different route costs in duarouter (trac #11)

[behrisch]

The estimated costs for a route should be identical for all vehicles. The counterexample can be found by executing the tools/assign/dua-iterate/simple test and looking at input_trips_10.rou.alt.xml.

Migrated from http://sumo.dlr.de/ticket/11

{
    "status": "closed", 
    "changetime": "2009-01-28T15:23:03Z", 
    "description": "The estimated costs for a route should be identical for all vehicles. The counterexample can be found by executing the tools/assign/dua-iterate/simple test and looking at input_trips_10.rou.alt.xml.", 
    "reporter": "behrisch", 
    "cc": "", 
    "resolution": "fixed", 
    "_ts": "1233156183000000", 
    "component": "routing (DUAROUTER)", 
    "summary": "different route costs in duarouter", 
    "priority": "major", 
    "keywords": "", 
    "time": "2009-01-26T13:37:49Z", 
    "milestone": "0.10.2", 
    "owner": "behrisch", 
    "type": "defect"
}

[behrisch]

• @behrisch changed resolution from "" to "fixed"
• @behrisch changed status from "new" to "closed"

[behrisch]

@behrisch commented:

The route cost calculation should probably be revisited. It is strange that the new cost of a route sometimes depends on historical costs and sometimes not. The Gawron diss says this should only happen if the route was really driven in the last iteration but it is not clear whether the implementation follows this requirement.