Arguments for locking at L3
The idea behind the locking was to limit the number of PB's and WT's as much as possible. So an Area Manager could check the non-drivable segments. If the AM thought they should stay, the AM would lock on L3 or otherwise delete the segment. If the segment was recently added the editor who added it should be contacted to explain why the segment is deleted.
The locking standard makes sense because of the routing consequences of improper mapped non-drivable roads with possible destinations. Disconnecting of a WT can have a great impact on the routing. Since the introduction of parking place suggestions based on walking time, connecting PB's is also of the utmost importance. Locking is important because the role of the non-drivable roads is not for every (starting) editor clear.
- In practice this only applies to the Netherlands, Belgium does not do this.
Arguments against locking at L3
- Locking at level 3 will block new editors from adding missing non-drivable segments or adjusting connected local roads, causing frustrations
- Since the introduction of phantom nodes, connecting non-drivable segments no longer creates unwanted nodes on the drivable segments
- There are hardly any reports of errors based on this sort of routing, making a high level lock feel like overkill
- House numbers are a lot of work, so the more editors can do this, the better it will be
- What the best place to park/stop with a car is, requires local knowledge, of course combined with editor knowledge. But again to more editors can edit these segments, the more local knowledge is used
Things everybody seems to agree on
Complex situations should always be locked appropriately. If a segment is important for proper routing, it should be protected against mistakes of starting editors.