make junction-blocking compliance (checkRewindLinkLanes) configurable (trac #1290)

[behrisch]

apparently in some circumstances (i.e. Italy) drivers just don't drive that way.
Even in Germany, compliance is not perfect

Also, VISSIM makes this configurable.

plan for implementation:

• add optional .nod.xml and .net.xml attribute keepClear. Defaults to "true". When set to "false", the junction blocking check will be disabled for that junction
• add netconvert opption --default.junctions.keep-clear which sets the default to "false" for all junctions when building a network.

Migrated from http://sumo.dlr.de/ticket/1290

{
    "status": "closed", 
    "changetime": "2015-04-29T08:00:02Z", 
    "description": "apparently in some circumstances (i.e. Italy) drivers just don't drive that way.\nEven in Germany, compliance is not perfect\n\nAlso, VISSIM makes this configurable.\n\n\nplan for implementation:\n* add optional .nod.xml and .net.xml attribute keepClear. Defaults to \"true\". When set to \"false\", the junction blocking check will be disabled for that junction\n* add netconvert opption --default.junctions.keep-clear which sets the default to \"false\" for all junctions when building a network.\n\n\n\n\n", 
    "reporter": "namdre", 
    "cc": "", 
    "resolution": "fixed", 
    "_ts": "1430294402895166", 
    "component": "microsim - junctions", 
    "summary": "make junction-blocking compliance (checkRewindLinkLanes) configurable", 
    "priority": "major", 
    "keywords": "", 
    "time": "2014-05-19T09:01:50Z", 
    "milestone": "1.0.0", 
    "owner": "namdre", 
    "type": "enhancement"
}

[behrisch]

• @behrisch commented:

add another junction attribute and a netconvert option to have it set by default

• @behrisch changed milestone from "0.21.0" to "1.0.0"
• @behrisch changed owner from "" to "namdre"
• @behrisch changed status from "new" to "assigned"

[behrisch]

@behrisch commented:

suggested attribute name: "keepClear"

[behrisch]

@namdre changed summary from "add option for disabling checkRewindLinkLanes" to "add option for disabling checkRewindLinkLanes (allow junction blocking)"

[behrisch]

• @namdre changed description from:

apparently in some circumstances (i.e. Italy) drivers just don't drive that way.

to:

apparently in some circumstances (i.e. Italy) drivers just don't drive that way.
Even in Germany, compliance is not perfect

Also, VISSIM makes this configurable.

• @namdre changed summary from "add option for disabling checkRewindLinkLanes (allow junction blocking)" to "make junction-blocking compliance (checkRewindLinkLanes) configurable"

[behrisch]

@namdre changed description from:

apparently in some circumstances (i.e. Italy) drivers just don't drive that way.
Even in Germany, compliance is not perfect

Also, VISSIM makes this configurable.

to:

apparently in some circumstances (i.e. Italy) drivers just don't drive that way.
Even in Germany, compliance is not perfect

Also, VISSIM makes this configurable.

plan for implementation:

• add optional .nod.xml and .net.xml attribute keepClear. Defaults to "true". When set to "false", the junction blocking check will be disabled for that junction
• add netconvert opption --default.junctions.keep-clear which sets the default to "false" for all junctions when building a network.

[behrisch]

• @namdre committed 78eacb7 (aka r18319):
  fix make junction-blocking compliance (checkRewindLinkLanes) configurable (trac #1290) #1290
• @namdre changed resolution from "" to "fixed"
• @namdre changed status from "assigned" to "closed"

[behrisch]

@namdre commented:

Via TraCI, drivers can already be "pushed" onto an intersection. Eventually there should be an non-TraCI way of configuring this behavior at driver level rather than network level. However this probably requires a whole new model for egoism/altruism/impatience/risk-taking/carelessness