So just to be clear, how I’ve mapped the lanes at this junction Permalink for turning left, and turning right across the dual carriageway, is correct?
That does indeed look correct, yes, though the lanes on the short segment aren’t strictly necessary in this case. There’s an argument for mapping them anyway, for completeness and consistency.
:? :?
Now I’ve been thrown off course again :lol:
Do you mean in the sense that lane guidance is overkill here for right turning traffic and isn’t needed? (I sort of agree as its difficult to end up in a wrong lane here, I only used this junction as an example). Or that I can delete the lane guidance for the short median segment, and the user will still get a “keep right in the right two lanes” instruction and visual guidance when turning right?
Yes, this. See this example for a working demonstration (Route Checker script will show lanes).
However, there’s no harm having the | ↱ ┆ ↱ | mapped on the median as well. (Sorry, I’m not helping am I?)
The user would only hear ‘Turn Right onto A6000 Chesterfield Road South’. They wouldn’t get any Keep right, lane information or visuals. Straight ahead lane guidance doesn’t show or create any voice instructions unless it is set to View Only, View & Hear or a TIO has been set on the node.
I disagree with this. In my testing, Waze is announcing the lanes along with the turns, even if they are ‘straight-ahead, Waze default’ on the previous node. As shown below:
This was in response to deleting the LG on the short median segment and only having LG on the first segment. As Atrophicshiner said LG will only show when a turn instruction is given so in that situation the user would only get a standard Turn Right instruction.
With LG on both segments the heuristics would announce the lanes.
I’ll try to find a suitable junction near me and do some drive throughs using the two different set ups to see which works, or if both work. Tile updates permitting, I’ll report back soon…
My example (link) above (and this one) proves that to not be true. You don’t need to put lanes on the median, unless you need to refine the incoming lanes further. The heuristics already takes this into account.
My quoted statement is still technically true as the right turn instruction is given :lol:
Can we not all agree that half mapping lane guidance through a junction is a bad habit to get into, are we not setting ourselves up for potential future problems should the junction mapping change, HQ tweaking heuristic behaviour, etc.
Does it really save us anything not mapping the median segment, and instead ensuring that lane guidance will work correctly every time.
How about we don’t map speed limits on segments shorter than 50 metres, that’ll save a couple of clicks and not many users will notice.
Looking at this junction: Permalink , this is a fairly common set up where you have slip roads off before you get to the # or H part. It doesn’t appear that heuristics work here. Ideally when approaching from the west, the user would see
↰ ↑↑↱↱
but I can’t see how that would be possible without simplifying the junction by deleting the slip segments.
You are correct, heuristics will only work on H or # junction and not include and slip roads in advance of the junction.
So simplification is needed to get LG to work fully on these junctions…but from some discussion on the Ireland channels, keeping it simple isn’t necessarily the way forward anymore. I’ll bide my time on these junctions then…
From the Wiki:
and this matches my on the road experience.
But I use metric, can anyone advise what happens when using imperial settings. Does LG appear converted exactly from metric into a strange number of feet, or is it rounded up to e.g. mile, 1/2 mile?
Dealing with a UR saying LG is confusing as it appears before another junction…
Can’t be 100% but I would say, based on evidence of other metric/imperial conversion in Waze, the app is unlikely to ‘convert’ the distance internally, instead it will trigger at the same location irrespective of the units shown in the app.
500m = 1,640.42ft
Depending on the location of the junctions, road types, etc, the first junction may need mapping or the 2nd removing, or something else entirely.
The junction/UR in question is here:
https://www.waze.com/en-GB/editor?env=row&lon=-1.22436&lat=53.12699&s=21059600383&zoom=5&segments=209226957
In metric land, the user would get an instruction around 300m out from this junction saying “In 1km, stay in the left/right lane for A38” In my opinion, thats not confusing and its user error if they turn left or right 700m too soon.
I just wanted to get an idea of when they get an instruction in imperial land. I’m on the road tomorrow, so I’ll switch Waze to furlongs and chains to see what happens :mrgreen:
Just being nosey…
Does the fact that the right turn is disallowed on the preceding junction have any impact on lane guidance instructions? In other words, if that turn into Amazon was allowed, would the instruction come after it, or does it still follow the 500m rule irrespective of other junctions?
For the record, the LG kicked in between 0.7 and 0.6 miles for major roads, and around 0.3 miles for smaller roads, so it seems like its using the 500m/1000m points rather than rounding up or down to a standard imperial unit (if such a think exists 
)
I think it follows the 500/1000 rule regardless. As I understand it, the issue the user has is: he wants to turn right onto the A38 at the T-junction further up the road. Waze tells him 1000m out from this junction that he needs to stay right in the right lanes and shows him the LG arrows. However 300m along from where he gets the instruction is a junction (with Hamilton Rd) where there is also a right turn lane, so he complains that this is confusing. I don’t agree though, if Waze tells you “in 1km use the right lane…” and then you turn right 300m later, Waze isn’t the problem…
Apologies, I had misunderstood this. I thought with the way the permalink was zoomed in around the Hamilton Rd junction that that was the intended lane guidance :roll:
I’d be inclined to agree with you, although it does depend on drivers’ perception of distance. There will become a point where the distance between is less obvious.
Also, FWIW, I would argue that this junction doesn’t need lane guidance. There is only 1 lane 300m out from this junction. One would hope that for anyone turning left they would not take the right lane at the point of when it splits into 2 as that is counterintuitive, and vice versa for anyone turning right. Clearly in this case it has caused confusion and the very nature of lane guidance is to reduce it.
There is also a second right turning into a PLR for which the user could confuse if they are not necessarily paying attention or familiar with the area/road names.
It feels like a lot of editors have rushed out to apply lane guidance and in a lot of cases they are not necessary IMO. Some, such as this one, actually go against the rules listed on Wazeopedia:
This is just one I came across and sorry to the editor who applied it - it’s not my intention to call you out