I've done since now work on the wayfinder revision draft, drawing mostly on previous comments by sketch and dwarflord.

I'm including a copy of the content here, the original text has been preserved for comparison in the first section.

The main changes are to the criteria, the naming guidelines, and the statement not to use "to" in the name.
======================

A wayfinder gives a user an instruction directing them how to continue on the road the user is already on, in situations where it may be unclear, and there would otherwise be no instruction. A wayfinder is warranted in the following cases:

• The continuation path for a controlled-access ramp, highway or freeway for an exit on the typical exit side, has at least two fewer travel lanes after the exit, than before it;
• The continuation path for an uncontrolled ramp or highway, or for an exit on the atypical side, has at least one fewer travel lanes after the exit than before it;
• The continuation path consists of a single lane; or
  Some or all of the associated signage may be interpreted as suggesting a split -- for example;
  • lacking typical exit language, or
    with arrows pointing left and right but no clear "straight-ahead" choice, or
  • is otherwise unclear or inconsistent so that drivers may be confused as to which way is the continuation or whether there is a continuation at all.

A travel lane is a lane which is long enough that a naïve driver might consider it a long-distance lane. For our purposes we will clarify this as any lane which;

• before nearing the exit, is neither striped nor signed as a departing lane over a significant length (approximately ¾ a mile, or its full length), and
• has not just started within a mile before the exit.

The original text of the wayfinder criteria to be replaced

A wayfinder gives a user an instruction reminding them to stay on the road the user is already on, in situations where a reminder is warranted. A wayfinder is warranted in the following situations:

• Lane drops, where at least as many lanes leave the road as stay on the road;
• Non-obvious continuations, where at least one "exit only" lane exists on the side of the road where exits are not normally placed (in a right-hand traffic country, exiting traffic is to the left and continuing traffic is to the right); and
• Inconsistent signage, where a highway continues as a numbered route, but signs call it only by a name.

In these cases, we need to use a wayfinder configuration. Each wayfinder configuration will have one "IN" segment and two "OUT" segments.

Keep these basic principles in mind:

• The OUT segments must be the same type (either Freeway or Ramp, as explained below).
• Each OUT segment must have a name different than the IN segment.
  • OUT segments should be named with the information displayed on the roadway signs.
  • However, if this would leave an OUT segment with the same name as the IN segment, the OUT segment should be left with no-name.

Previously in these situations segment naming was modified in one of the following ways in order of preference:

• Add a control/destination city, if it is known, to the OUT segment name.
• Remove the compass cardinal from the OUT segment name.
• Add a space to the end of the OUT segment name (note that this will be flagged by the WME Validator script, if active, and possibly result in undoing by another editor)
• Create a stub in the IN segment just before the junction with no street name.
  • NOTE: This option should only be used as a last resort because it prevents the routing server from properly calculating turn delays, and may result in inefficient routing through the area.

If you see any of these set-ups, please correct it to one of the two the proper methods listed above.

Do not use "to" at the beginning of the name of the segment if it is for a continuation of the same road, e.g. a freeway split where one side is the continuation of the same freeway (the user is already on the road, not going "to" it).

Each "stub" segment should be 5 m (16 feet) long. This is long enough so it will not cause routing problems, but it is short enough to suppress display of the names (on Freeway stubs) and keep freeways looking contiguous (on Ramp stubs).

The OUT segments should have a turn angle of 10°-20° on opposite sides of each other, consistent with the geometry for freeway splits. This will allow for easy identification in WME, but still look seamless in the client app.

To configure the wayfinder,
If the numbered/signed exit is on the right, the OUT segments should be Ramp type. This will give an "exit right" instruction for the exit and a "stay to the left" instruction for the continuation.

For a basic exit, use a named Ramp stub for the continuation, and a standard Ramp for the exit.

For a freeway split, use a Ramp stub on both sides.

If the numbered/signed exit is on the left, the OUT segments should be Freeway type. This will give "stay to the" instructions on both sides (if Ramp segments were used, Waze would give a confusing "exit right" instruction for the continuation).

For a basic exit, use a named Freeway stub for the continuation, and an unnamed Freeway stub followed by a named Ramp for the exit.

For a freeway split, use named Freeway stubs on both sides.

======

Please reply here if think any modifications are necessary or not, so we can finally have these changes incorporated into the wiki. Thanks to those who came before and fleshed most of this out.

CBenson wrote:This all sound good to me.

I wonder about this:

PesachZ wrote:The OUT segments should have a turn angle of 10°-20° on opposite sides of each other, consistent with the geometry for freeway splits. This will allow for easy identification in WME, but still look seamless in the client app.

I don't really see the need for the angle. But I can agree to this if necessary.

This for consistency with the rest of the interchanges article. And also provide the added benefit of making the turn restriction and report>closure arrows be easily selectable.
But looking at it now I think I well clarify that that sentence to read:

... The OUT segments should have a turn angle of 10-20° off the midline on opposite sides of each other, ...

This will mean that the junction angle between the two OUT segments will end up being between 20-40°.

Bigbear3764 wrote:Something to consider, and I'm not sure where you would put. I had someone on chat with a UR where someone was routed off and back on a highway. When I started looking at the road names to make sure the detour prevention mechanism was working properly, I found a segment with no name. The person on chat said he had just requested a wayfinder at the interchange. So in order to get the 2 out Freeway segments to be different names, he had changed the one to No Name.

It was a simple fix to divide the No Name segment and to name it before the next ramp to get the detour prevention mech to work.

Maybe just a note to be aware of other situations in the area like Detour Prevention.

This guideline is only for the first 5m of the segment to be a no name stub, after that the road continues with the property segment names. This was designed to preserve detour prevention mechanisms. Although a note can be added to that effect so editors understand why a stub is better that an entire no name length

kentsmith9 wrote:We need to be clear in the Wiki how to interpret the "To..." element
In this location of Fremont, CA, Mission Blvd is a short 1 mile segment that links between I-880 and I-680. From I-880 you see:

To_I-680.png

1. "Mission Blvd" is the road you are about to exit upon. It also happens to be SR-238, but this sign does not note that.
2. "Sacramento" not the target city for Mission Blvd, but it is actually for I-680 N.
3. "To I-680" is one mile up Mission Blvd.

In this case the "To..." is actually related to both I-680 and Sacramento. If you stay on Mission Blvd, you end up in Oakland, not Sacramento.

I currently have the exit for this segment entered as "Exit 12A: Mission Blvd / to I-680 / Sacramento"

I would agree in that case, and it follows the naming principle (which I read in the wiki, but can no longer find right now) to match the BGS right to left, top to bottom.

I think we all agree, and I will update the draft for the 'to" statement, that a valid exception is when the BGS says "to...". Though this is a very rare case, as the statement only applies to the continuation party of the wayfinder, and these example given here are all for the exiting side.

The statement not to use " to..." Was only made in regards to a continuation on the same road, not an exit ramp. Nevertheless, since there is the possibility that a highway continuation may have "to..." on the wayfinder BGS, (usually very near the termination of a highway) I will make this adjustment to the draft.

KB_Steveo wrote:If we are also editing the JSG/Intersections page, I would like to propose including leaving the segment between 2 interchange roundabouts unnamed.
http://i1276.photobucket.com/albums/y479/sbelekevich/0e5c8014-4a90-4bd2-95c8-977dc5b73479_zps1faefb9c.jpg

This provides the following benefit:
When you are unfamiliar with the area and you see this after passing through the east most roundabout (at the blue arrow):http://i1276.photobucket.com/albums/y479/sbelekevich/MasonSt1_zps776ddcdd.jpg if you are supposed to go on US-41 S,

• You hear: "go straight, then turn left to US-41 S (paraphrasing)
• You see US-41 S at the top of the client, giving you a visual cue along with the verbal that you probably want to be in the left lane at the first roundabout (there is another 3 lane choice at the next).

If you are going straight, you see Mason St at the top, with the same benefit.

I remember my first time driving through this when it opened, and it was definitely intimidating. I know this is kind of a specific type case, but it is happening all over Wisconsin, and I am constantly catching other editors naming the middle segment thus removing the 'next step' visual cue. Although this is more suited for larger interchanges, it still works at smaller interchanges since the signage is still there, even though it may not be "necessary".

I can't be sure, but I don't belive the "and then" next step instruction is controlled by whether or not that middle segment is named. I think it has to do with the length (of time or distance) between the two turns. I'm sure someone will correct me if I'm wrong though

PesachZ wrote:

kentsmith9 wrote:We need to be clear in the Wiki how to interpret the "To..." element
In this location of Fremont, CA, Mission Blvd is a short 1 mile segment that links between I-880 and I-680. From I-880 you see:

To_I-680.png

1. "Mission Blvd" is the road you are about to exit upon. It also happens to be SR-238, but this sign does not note that.
2. "Sacramento" not the target city for Mission Blvd, but it is actually for I-680 N.
3. "To I-680" is one mile up Mission Blvd.

In this case the "To..." is actually related to both I-680 and Sacramento. If you stay on Mission Blvd, you end up in Oakland, not Sacramento.

I currently have the exit for this segment entered as "Exit 12A: Mission Blvd / to I-680 / Sacramento"

I would agree in that case, and it follows the naming principle (which I read in the wiki, but can no longer find right now) to match the BGS right to left, top to bottom.

I think we all agree, and I will update the draft for the 'to" statement, that a valid exception is when the BGS says "to...". Though this is a very rare case, as the statement only applies to the continuation party of the wayfinder, and these example given here are all for the exiting side.

The statement not to use " to..." Was only made in regards to a continuation on the same road, not an exit ramp. Nevertheless, since there is the possibility that a highway continuation may have "to..." on the wayfinder BGS, (usually very near the termination of a highway) I will make this adjustment to the draft.

I made this adjustment and a bunch more to the draft page for the wayfinder section, please have a look there.

To summarize;

• Clarified that a wayfinder can be any (primary) road type. I only mentioned the primary roads, since this is on the interchanges subpage, on the regular intersection page it can mention the non-primary road types.
• Added exceptions to the 'don't use to' rule for if to is on the BGS, or if the name on the BGS is not the name of the continuation rd.
• If "To" is on the BGS put it last in the name, unless the control city only applies to that road, then put the control after the "to" highway.
• If the continuation rd is not first on the BS it should nevertheless still be first on the segment name?
• added in images of sample BGSs for these exceptions.

DwarfLord wrote:

ottonomy wrote:

sketch wrote:Oh, also, I'm not crazy about the "2 dropped lanes on the expected side" rule, exactly—what of the situations with 1 continuing lane, 1 option lane, and 1 dropped lane? The "only 1 lane continues" rule doesn't really apply unless we explicitly exclude option lanes from that count.

Obviously, we do need to have some sort of objective rules for this spelled out in the wiki, but this is a subject where any number of factors could make exceptions to the rules seem reasonable.

There are places where only one existing travel lane peels off to another route, but there's little advance signage, poor road striping, and/or maybe a blind curve, which combine to make the exit sneak up on drivers.

There are other places, where five lanes split into 3 and 3 (option lane in middle), but the BGS have been warning of the split every half mile for 5 miles or so, and giant Y arrows loom above, ominously, sign after sign.

The former situation begs for a wayfinder. The latter should obviously have one, but almost doesn't need it.

Some editor discretion will have to come into play from time to time.

Absolutely! Original intent gave high value to what could be called the "discretion" condition:

Some or all of the associated signage or striping, or the geometry of the road itself, may be interpreted as suggesting a split -- for example:

• some or all signs lacking typical exit language;
• signs with arrows pointing left and right but no clear "straight-ahead" choice; or
• otherwise unclear or inconsistent signage and striping so that drivers may be confused as to which way is the continuation or whether there is a continuation at all.

If readers see that this "discretion" condition is a significant part of the guidance -- maybe it could simply be placed at the front of the list rather than at the end? -- it could accommodate many situations where a purely quantitative approach would be inadequate.

[EDIT: broadened suggested "discretion" language.]

I changed the text of that last condition to read

Some or all of the associated signage may be interpreted as suggesting a split (use your better judgement, and discretion) -- for example;

DwarfLord wrote:

PesachZ wrote:I changed the text of that last condition to read

Some or all of the associated signage may be interpreted as suggesting a split (use your better judgement, and discretion) -- for example;

I've reworded the "discretion" item of the warranted-wayfinder list to read:

Some or all of the associated signage or striping, or the geometry of the road itself, may suggest a split to drivers (use your better judgement, and discretion) -- for example:

• some or all signage or striping lacks expected "exit" indications;
• signs are present with arrows pointing left and right but no sign clearly establishes the continuation; or
• signs and striping are otherwise unclear or inconsistent so that drivers may be confused as to whether they are encountering an exit or a split.

Hope this meets with approval, otherwise it's fine to revert or further modify.

You mentioned the geometry of the road, but all of your examples are talking about the signs only.

I added more example images to the draft wiki page, and structure. and more definition when (not) to use "to " in the name.

DwarfLord wrote:My understanding is that it's not a question of legality but rather whether there is ever a situation where one would actually want Waze to route off a freeway only to get right back on it again. If not, why not leave it red. Cheap insurance!

Yes there a good reason occasionally. As a matter of fact, sketch mentioned two already [emoji1]

sketch wrote:Good catch. I think that guidance should be reversed. Let detour prevention do its job. Enabling the thru movement allows people who mistakenly get off to be told to get back on again, and perhaps more importantly allows for closures on freeway segments with detours that use interchange ramps—which is common—without having to wait for a tile update to enable that movement.

The second reason will also help for instant closures reported via the app for say an emergency closure after a motor vehicle collision, flooding, etc.