People: Map Problems <> Update Requests

When discussing, let's try to be precise about what we're seeing and talking about so that there isn't confusion.

Recently, the line between system-generated and user-submitted have been blurred because Waze is testing out some new processes in identifying things like incorrect city name (Portola Valley for example) which are being fixed automatically, but if it isn't being fixed, then the script is posting Update Requests (normally reserved for end-user submissions from the app) with the notes starting with "Waze Automatic:"

Update Requests from real users will often include a purple and green line which is relevant only to that user/submitter. Map Problems will show purple/green lines which are aggregated from many drives.

Map Problems identification code runs regularly and analyzes somewhat old data.

The "Suggested route frequently ignored" appears to be the type of Map Problem which has appeared as an aging (orange already) Map Problem today. At least, that is what I'm seeing in my area. And they do, as noted in the earlier thread, disappear often when zooming.

shawndoc wrote:Its too bad no one has invented any sort of testing platform, where Waze can try out scripts and other changes before rolling them out to the production servers.

They have one, but seemingly choose not to use it for all changes.

It seems that if there are under a certain number visible, that it works ok after pans. But once I get A LOT on the screen at once, they disappear. What's really odd is that the "normal" Map Problems appear just fine regardless, and only all these "didn't follow route" ones appear to be affected by the display issue.

It's been mentioned how old this data is, but I just found one that appears to be analyzing data almost a year old.

https://www.waze.com/editor/?zoom=7&lat ... s=33798653 At that junction, the Waze recommended route is to turn left, against the 1-way segment. This segment hasn't been updated since March 2012, which would be the most recent date it could have been set incorrectly and allowed traffic the other way. Now, true enough, Waze does route against 1-way segments, but there would need to have been significant other errors nearby to cause this.

This makes me dubious that researching/reveiwing any of the recent "sea of orange" map problems is worth anyone's time...

CBenson wrote:There are also quite a few where I can't make heads nor tails about what the green line is representing.

As it goes along, especially at further distances from the MP, it appears to be "averaging" the drives. Such as something like the OP in this thread: http://www.waze.com/forum/viewtopic.php ... ge#p160526

shawndoc wrote:Hey Champs, any word from on-high about when the Problem Layer will start working correctly again?

Nothing back yet aside from a confirmation that the know it's a problem & have passed it to the developers...

jemay wrote:The sea of orange is back.... But they appear and then disappear and several screens refresh reappear. Has anybody else noticed this also this morning?

A lot of them are "Turn Problem: The displayed route is frequently not taken by users"

Yes, but it's a different "sea of orange" problem... See viewtopic.php?f=8&t=37502

I'm killing them as fast as I can. My finger is cramping...giving the screen the permanent bird.

Finally, got L.A. cleared up. My poor mouse got clicked almost to death...