Splits? Splitting a way leaves one way with the history and creates a new way. There's multiple changes in lane count, speed limit, advised speed, and more that would have necessitated splits.

It wasn't deleted and redrawn, it may have crept in from a join and retag since I'm not usually looking for or expecting layer=1, 0 or -1 in places where the layer=* tag isn't necessary because the implicit values for bridge=yes, tunnel=yes or an absence of both are sufficient.

The theoretical gore is where I placed the end of the transition on I 405. Sorry, can't use Google as a citation for map data, not only are they usually wrong, they're also copyrighted and using it would be a copyright violation.

The theoretical gore does NOT start at the end of the solid line as you're suggesting. The solid line is why the segment has a change:lanes=no|not_left|yes|... tag. If Tesla isn't able to handle that, then the problem isn't that it's not mapped accurately, it's that Tesla is not correctly interpreting the data.

OK, that's not where the ramp should start. The ramp should start where the painted gore starts. That's handled by the change:lanes tag. See way/781574733

Adding turn lanes for direction confirmation arrows on offramps aren't useful, it's the actual function of the lanes that is.

Please keep an eye out for placement tags, which provide a hint where the line needs to go. This edit resulted in lanes on the ramp being shifted about half a lane right of where they're actually located.

Please keep an eye out for lane and placement tags, as this will affect where the line itself needs to be placed. You moved all the lanes on this ramp out of alignment to the right.

This changeset has been reverted fully or in part by changeset/82221598 where the changeset comment is: See previous note.

name is not ref, don't use refs in the name tag. You also added embankment tags where they don't apply, and extended turn lanes past their exits. This edit makes very little sense.

Removing hov=designated isn't correct, as this leaves just access=no as the access tag, which means nobody can use that road. access=no, hov=designated means only carpools can use that road.

This changeset has been reverted fully or in part by changeset/82221199 where the changeset comment is: See previous note.

Thanks for the update on it not being an exit number, however, extending the ramp long past its split and erasing lane guidance is not the way to go around this. If your autopilot is failing to get this right when the map is matching ground truth, you need to contact your manfuacturer and report a bug. See also: osm.wiki/Tagging_for_the_renderer

I agree the wording is a little odd on the challenge, however, aeroway=helipad is literally just the helipad.
---
#REVIEWED_BAD #OSMCHA
Published using OSMCha: https://osmcha.org/changesets/82190473

Yeah, pretty sure a helicopter landing there would punch straight through the roof. Map the helipad where it is.
---
#REVIEWED_BAD #OSMCHA
Published using OSMCha: https://osmcha.org/changesets/82190473

Might check against the GPX traces in the area to line up the background to known roads. I think Bing and Maxar are closer last time I was mapping in that part of Oklahoma.

Not wholly sure where you're getting a Y where it looks to be more of a T.
---

Published using OSMCha: https://osmcha.org/changesets/82161488

This geometry feels especially wonky.
---

Published using OSMCha: https://osmcha.org/changesets/82161596

Doing pretty good, but might want to doublecheck your imagery alignment, I feel like everything's shifted northwest a few meters.
---

Published using OSMCha: https://osmcha.org/changesets/82162636

Pools should be mapped as pools, not natural bodies of water.
---

Published using OSMCha: https://osmcha.org/changesets/82168204

This changeset has been reverted fully or in part by changeset/82184305 where the changeset comment is: Still under construction.