Default turns when adding new roads

[txemt]

You'd be surprised how quickly you could clean that up with QW.

[Kuhlkatz]

You'd be surprised how quickly you could clean that up with QW.

+1

The alternative to that sea of red is a sea of orange.....
Just imagining each of those as a UR or a MP should be enough encouragement to spend 5 minutes tops doing a Q/W on each node and giving the area a quick once over to see if road markings includes one way arrows somewhere....

Sent from my HTC Sensation Z710e using Tapatalk 2

[CBenson]

You'd be surprised how quickly you could clean that up with QW.

+1

The alternative to that sea of red is a sea of orange.....
Just imagining each of those as a UR or a MP should be enough encouragement to spend 5 minutes tops doing a Q/W on each node and giving the area a quick once over to see if road markings includes one way arrows somewhere....

Sent from my HTC Sensation Z710e using Tapatalk 2

-1
I've got to disagree here. A quick once over to see if road markings includes one way arrows somewhere before mass QWing is much more likely to produce URs and MPs than the existing soft turn restrictions. Unless of course you get lucky and there are in fact no one-ways or other turn restrictions there.
If these soft restrictions would cause a sea of orange, then that would have attracted editors to correct them, which just hasn't happened. (Yes, there are active editors that monitor this area.)
I'm not saying they shouldn't be made correct hard restrictions. I'm just saying that they should be done correctly.

[Kuhlkatz]

I've got to disagree here. A quick once over to see if road markings includes one way arrows somewhere before mass QWing is much more likely to produce URs and MPs than the existing soft turn restrictions. Unless of course you get lucky and there are in fact no one-ways or other turn restrictions there.

Someone familiar with the area should be able to correct that fairly quickly. In most populated neighborhoods, one-ways and fancy turn restrictions will not exist apart from the main arteries through them. If it's near a CBD or an industrial area, it may be a different story.

I understand that the auto-directionality and soft turn restriction overrides should not be happening any more in the back end, even though there has been isolated recent reports where this still seemed the case.

It still does not make sense why soft turns are still allowed by default when a junction is created, yet the default for the 'hard' restrictions still deny everything.
A few trips through there would maybe not affect routing due to the soft settings, but with the new automated MP system, each junction traveled through a few times will potentially create an auto-MP as the hard restrictions do not match.

[CBenson]

Someone familiar with the area should be able to correct that fairly quickly.

Yes, that is correct. But as Kodi75 stated its a matter of priority.

In most populated neighborhoods, one-ways and fancy turn restrictions will not exist apart from the main arteries through them.

Well, "will not exist" is a fairly strong statement. I've got plenty of residential areas where there are a few one ways or other odd restrictions that are not on the main arteries.

If it's near a CBD or an industrial area, it may be a different story.

Well that example is within the 24th largest city in the United States, but its not downtown or overly industrial.

I understand that the auto-directionality and soft turn restriction overrides should not be happening any more in the back end, even though there has been isolated recent reports where this still seemed the case.

Did I miss something? I thought that waze did still automatically override soft restrictions (including directionality such as unknown).

It still does not make sense why soft turns are still allowed by default when a junction is created, yet the default for the 'hard' restrictions still deny everything.

I'm missing something again. What's a "default" hard restriction. I thought hard restrictions were by definition manually created (although frequently unintentionally by editing a segment property without properly setting the restrictions).

A few trips through there would maybe not affect routing due to the soft settings, but with the new automated MP system, each junction traveled through a few times will potentially create an auto-MP as the hard restrictions do not match.

Sure, but now we are back to the priority issue. Should that happen, the area will get editor attention and these restrictions will be cleaned up.

[Kuhlkatz]

Did I miss something? I thought that waze did still automatically override soft restrictions (including directionality such as unknown).

I was referring to one thread that included Tony's comment as below. Not sure that it was this particular one, but I do find this interesting.

http://www.waze.com/forum/viewtopic.php?f=10&t=36644&p=340911#p340911 :

on Tue Jan 15, 2013 6:53 pm

wrote:
No. I've had a street go from two-way to one-way, simply because I was the only person driving down it, and I always drove in the same direction.

This is no longer the case as of about a month or so ago. Prior to that it was possible for *some* segments to still be under automatic direction updating even after having been manually edited. This now no longer happens once the segment has been edited again.

To reiterate, as of about a month ago, editing *any* property of a segment will now cause direction on that segment to become fixed.

In the same thread gerben also said :

* Unfortunately, during the infrastructure change a lot of roads did not get this 'internal directionality lock' when edited. This was fixed somewhere in december, so only for very recently edited roads you can be sure that the directionality will not be updated by any automatic process.

So based on this, if a new area is layed out and the names added as a seperate pass, or as I prefer, lay out and name the full-length streets first and then connect them afterwards, can we assume they were edited at least once more after the initial creation, so technically speaking their directionality *would* remain fixed, even if it is incorrect.

I'm not disputing anything or trying to pick fights here, I'm just curious about why some of you first-generation editors (read "old hands at this") prefer to leave sections like these, when a QW clean-up is fairly quick, especially if someone knows the area.
I personally prefer to QW junctions while I create a section, so I know it's 'fixed'. Afterwards, red arrows when viewed with 'Connectivity Arrows' (Shift-Z) draws attention when zoomed out, and can be quickly verified as a valid turn restriction or not. To spot an invalid turn restriction here is going to be impossible if not cleaned up.

I thought hard restrictions were by definition manually created (although frequently unintentionally by editing a segment property without properly setting the restrictions).

So, as far as turn restrictions go, if they are never explicitly set using Q/W on the junction or clicking any arrows with the segment highlighted, only the allowed 'soft' ones are in place and no hard restrictions should exist for the segment / junction pair ?

What I am trying to get at is this : Do we know for sure that sections left like this would not generate auto-MPs due to intact 'soft' restrictions allowing everything, and having no other explicit restrictions set, or would it be better to spend a few minutes and take the precautionary approach ?

I assume someone would have been swamped with MPs, but it more than likely would require regular drives over the area as well to provide any results that could generate MPs.

[CBenson]

So based on this, if a new area is layed out and the names added as a seperate pass, or as I prefer, lay out and name the full-length streets first and then connect them afterwards, can we assume they were edited at least once more after the initial creation, so technically speaking their directionality *would* remain fixed, even if it is incorrect.

I'm getting in over my head here. The problem is that my experience is with base mapped roads and you are assuming a situation where an editor has laid out the area. So yes I think you are correct. But there are plenty of segments around here (although admittedly fewer all the time) that have never been touched by an editor at all as they were base map imports.

I'm not disputing anything or trying to pick fights here, I'm just curious about why some of you first-generation editors (read "old hands at this") prefer to leave sections like these, when a QW clean-up is fairly quick, especially if someone knows the area.

I don't prefer to leave sections like these. But a careful review of which turns should actually be restricted should be a part of the clean up. I just object to those that QW everything without taking the time to put back the turn restrictions that actually should be there.

I personally prefer to QW junctions while I create a section, so I know it's 'fixed'. Afterwards, red arrows when viewed with 'Connectivity Arrows' (Shift-Z) draws attention when zoomed out, and can be quickly verified as a valid turn restriction or not. To spot an invalid turn restriction here is going to be impossible if not cleaned up.

Right, when adding a new communities, which we don't get to do too often here, that's good practice.

So, as far as turn restrictions go, if they are never explicitly set using Q/W on the junction or clicking any arrows with the segment highlighted, only the allowed 'soft' ones are in place and no hard restrictions should exist for the segment / junction pair ?

That sounds right, but again most soft turn restrictions that we are dealing with here are base map imported roads that have never been touched by an editor.

What I am trying to get at is this : Do we know for sure that sections left like this would not generate auto-MPs due to intact 'soft' restrictions allowing everything, and having no other explicit restrictions set, or would it be better to spend a few minutes and take the precautionary approach ?

In the absence of anything else to correct on the map, it would be better to spend the few minutes and eliminate the soft turn restrictions. But if there are other URs and MPs around, I'm going to be addressing those areas rather than searching out soft turn restrictions in residential areas. The point is not that the soft turn restrictions should be left in place. The point is when converting them to hard restriction you should be reasonably sure that the turn restrictions are correct. You should not blindly QW an area that is full of soft turn restrictions. I think we are in agreement here as long as we recognize a clear difference between "blindly QWing" and "giving the area a quick once over."

[Kuhlkatz]

The problem is that my experience is with base mapped roads and you are assuming a situation where an editor has laid out the area. So yes I think you are correct. But there are plenty of segments around here (although admittedly fewer all the time) that have never been touched by an editor at all as they were base map imports.

Ah, that provided the 'Lightbulb moment' -
Being basemapped, NA likely had most of the map looking this way initially. I was referring to and also 'assuming' that it was a newly laid out area, seeing that it was already named etc.

I tend to forget that not everyone started with a blank slate, as I am only relating to what I see day to day. In INTL server case, except where basemapped, that would boil down to sloppy editing 99% of the time. In your case that's the norm for untouched areas, and unfortunately the first poor sod that drive those segments are blamed for the mess, as it now contains his name