Here's an example I did nearby similar to the box diagonals method Ottoman talks about.

It's a tricky setup, but done right generally mitigates a lot of the routing and timing issues inherent to 'standard' box intersections.

I drive the Old Baltimore Pike/SR-896 intersection daily, so I'll keep an eye on routing/timing and the like when we start testing out this idea.

Whoops, sorry about that! Here's the link: https://www.waze.com/editor/?env=usa&lo ... s=79735655

And I think Ottobox is hilarious. Do it!

I don't know what it looked like before so I can't make any judgments on what was there, but we definitely still need the ability to allow u-turns on Route 7 here. Can someone tackle that?

mishterhaz wrote:Whoops, sorry about that! Here's the link: https://www.waze.com/editor/?env=usa&lo ... s=79735655

And I think Ottobox is hilarious. Do it!

MReiser4670 wrote:I don't know what it looked like before so I can't make any judgments on what was there, but we definitely still need the ability to allow u-turns on Route 7 here. Can someone tackle that?

I fouled up by clicking the link above to see the intersection being discussed. My bad!

However, with that being said, the intersection of US-40 & SR-7 doesn't allow u-turns for either direction of SR-7. That's the one I saw needed u-turns added. Anyone here disagree?

russblau wrote:It seems like we would need to add a junction between these two segments, and likewise on the other diagonal corner. Any problem with that?

Looks correct to me, that seems to be the best way to go about it. It would be a shame to mess up such an otherwise nice intersection!

viewtopic.php?f=622&t=122844

I've been seeing a lot of map errors for disconnected roads that indicate that they are from missing u-turns. Regardless, that H-junction is too wide to stop a u-turn so a specific AGC is not really needed so I don't think that applies here.

What should be done about this one? If it's to stay, it should be a MH not street. And it really shouldn't have a street name of "u-turn" on it.

I don't think it should stay just to get around a limitation of the lack of u-turn instruction in Waze, else we'll be putting these things everywhere. In other locations, if we need to draw a u-turn to fix a map problem, probably should look at if stretching the junction out further than 15 meters so a u-turn is allowed will work first.

Well in Delaware I'm just a contributing editor, so it's y'alls decision! I'm just here to help!

btw, I Iike how your sig is formatted. Hope you don't mind if I am inspired by it!

This still bothers my OCD. For traffic timing stats, a left-turn and u-turn out of that node is going to take the same amount of time so dropping that AGC in there is going to defeat that. Before when traffic stopped at the light, the amount of time it took to do the left turn at that intersection vs going straight was kept track of at the node that crosses Old Baltimore Pike. Now that's not possible since there's a node at the stop line further back for the u-turn. So it can only track traffic going either to the left through the u-turn or "straight." Left turn traffic has to go straight through that node and then left at the node at Old Baltimore Pike. So we just made Waze less accurate at predicting traffic timings there.

The only place for timing purposes an AGC should be put in is for the two lefts off of Old Baltimore Pike since people wait at the light behind the one node, then proceed straight through the node and then left at the second node so it doesn't really track the traffic delay properly for left turns there.

Yeah, I'm pontificating like I'm an expert because this is a something I just learned about during a recent map raid when Ottonomy went over how this stuff works, and I was like, "woah." I was so impressed I copied his chat text to my wiki page.....

https://wiki.waze.com/wiki/User:Poncewa ... _Junctions

I PM'ed PrettyInterstateMachine and got a fairly compelling response back about the early break for the u-turn -- but I would say that also applies to the left-turn there since the stop line is back that far too.

I'd like to try to modify this intersection a bit to address all of our concerns, and the original editor's issue. Russ, could you lower these segments to 3 so I can work on it please?

Permalink

I'm going to model it a bit after that chat Ottoman gave on how to handle box intersections, even though this really isn't a full box.

Thanks.