Time for an update - and we are now pretty confident that the problem is resolved.
Our resident genius, Timbones
, has been monitoring several cross-London routes for a few weeks by placing routing requests through the livemap and recording the results. They've pretty much all shown the same pattern, with a failure to return routes during morning and evening rush hours.
Below is the graph for the last 5 days on a route from Cheshunt to Wandsworth
As a London driver, I would agree with the routing choices: during peak hours, the North Circular can be pretty jammed, so the preferred route tends to be via the M25. As traffic lets up, particularly late at night, the North Circular is the better choice.
You can see the tail end of the failure period for Wednesday evening. The fix was implemented Thursday morning (we think
that's why the routing servers were down for a while). Since then, routing has been almost entirely successful at all times of day.
We (the Country Admins) have speculated endlessly on the cause of the problem and the nature of the fix. I'm sure you have as well, but we don't have any answers. It's understandable that the routing algorithms are part of Waze's proprietary intellectual property, so we don't really expect to be told much detail.
What I can tell you is that this is a problem that hasn't just affected London. It is a problem with the routing and is not due to any particular map problem. What Waze have currently done is implemented a change to work around this problem rather than change the routing algorithms, so there may be further developments at some point in the future.
I'd like to thank you all for your patience with this issue. Please remember that most of the work on pushing this issue was done by dknight212
and I've only taken over in time to pick up the glory!
I'm sure he'll pop by to see what's happening.
If you have any feedback, please continue to post in this thread - whether you are still seeing routing problems across London, or, indeed, have experienced a marked improvement.