Get a sneak peek at whats next for Permanent Hazards on our April 7th Office Hours!
You MUST read the "[Guidelines] How to request and search for new features" before posting here.

If you think you've found a BUG, use the appropriate WME version release thread found in the Waze Map Editor forum (parent of this forum).

Post Reply
Forum rules
You MUST read the "[Guidelines] How to request and search for new features" before posting here.

If you think you've found a BUG, use the appropriate "Official Feedback" thread found in the Waze Map Editor forum (parent of this forum).

[Protecting Maps] Subsystem maintainer like approval process

Post by bemk92
Hi,

I've recently found waze and am quite interested in editing the map.

With my new view, I found a possible area of improvement.

The behaviour I'm currently observing, is that in order to get a change to be made at all, an editor can only touch roads that are its level, or lower.
That means a rank 1 editor can edit all roads at rank 1, and potentially harm those roads.
Also, edits to higher ranked roads can not be made, when a new situation has recently been formed.

The current solutions to this are either a rank change request, which then lowers the rank required to edit a road (which then potentially allows other lower ranked editors to damage that section of road) or to have a higher ranked editor make the change for you (which damages the lower ranked editor in its points).

The system I'm proposing is slightly different and is much like the Linux kernel development model.

In my proposal, I'd suggest allowing edits to be made to any road by all editors (regardless of rank). However, when the time for applying those edits comes, rather than immediately applying the edits, it would be wise to send the proposed change to an editor with appropriate permissions for review.

The advantages of this system would be as follows:
1) Lower ranked editors would be able to make a more meaningful contribution
2) Road ranks wouldn't have to be lowered in order to allow lower ranked editors to edit (enhances safety)
3) Lower ranked editors would be barred from making harmful edits.
4) Lower ranked editors can still notify higher level editors without making a report on site (after all, whilst driving we need the eyes on the road and the hands on the wheel).

The disadvantage would be:
1) Higher ranked editors would potentially see a whole bunch of edits up for review by rank 1 editors.

As an additional side feature, the higher level editors would get points for allowing or denying these edit reviews, in order to encourage the processing of edits up for review, in turn lowering the number of edits up for review.
bemk92
Posts: 2
Has thanked: 1 time
Been thanked: 2 times

POSTER_ID:17017902

1

Send a message

Post by irowiki
Hey, this is close enough to what I was thinking, so I'll just bump it.

I was thinking the recent issues of editors and lock levels and roads being deleted and such could be easily solved by doing something like in the OP, but with a twist. The code for approving places is already in the editor, so why not extend that to edits themselves?

Like when wanting to connect a PLR to a level 2 road and you are a level 1, it'll let you connect the road and set restrictions but then it goes into an approval state, where a higher level editor will have to toggle approve/reject.

Level 4 and you delete a MH? A level 5 would have to approve it. A level 5 goes on a rampage and deletes every interstate he can? Well a level 6 (or waze staff) by that point would need to approve it. (or maybe another 5?)

The downside is this could potentially be a LOT of approving needed, so maybe it would just be for deleted segments.

The upside is it would streamline unlock requests perhaps.

I have much more in my head regarding this idea but I'll leave it here for now.
irowiki
Posts: 250
Has thanked: 66 times
Been thanked: 37 times
Send a message

Post by irowiki
I don't think Map Maker has the community we have, though. I have had some very bad run ins with Map Maker's "moderators" as well.

Like people 1,500 miles away rejecting edits that I'm making to my own home town based off of street view or sat view that is years out of date.
irowiki
Posts: 250
Has thanked: 66 times
Been thanked: 37 times
Send a message

Post by irowiki
With the trend in lock levels getting higher, low level editors can't edit anyway, so how do we solve that? Would be nice to have some sort of middle ground, even letting out of area high level editors approve edits, for instance.
irowiki
Posts: 250
Has thanked: 66 times
Been thanked: 37 times
Send a message

Post by t0cableguy
I think this is moving towards Google Map Maker's flawed system. The updates are provided.. and sit there for months waiting on approval.
t0cableguy
Posts: 1082
Has thanked: 306 times
Been thanked: 412 times
Send a message
I'm no longer editing. Please don't contact me with things regarding Waze, unless you want the info on how to take over the SE URcomments list. Then email or send me a message on GHO at gmail.com t0cableguy@

Post by t0cableguy
But that is the path to the dark side lol
t0cableguy
Posts: 1082
Has thanked: 306 times
Been thanked: 412 times
Send a message
I'm no longer editing. Please don't contact me with things regarding Waze, unless you want the info on how to take over the SE URcomments list. Then email or send me a message on GHO at gmail.com t0cableguy@

Post by t0cableguy
the trusted user system is flawed because we have IGN editors approving bad place additions, and the gps points are way off for where the places are added, especially when the users are adding them from the highway in their car.
t0cableguy
Posts: 1082
Has thanked: 306 times
Been thanked: 412 times
Send a message
I'm no longer editing. Please don't contact me with things regarding Waze, unless you want the info on how to take over the SE URcomments list. Then email or send me a message on GHO at gmail.com t0cableguy@

Post by taco909
t0cableguy wrote:I think this is moving towards Google Map Maker's flawed system. The updates are provided.. and sit there for months waiting on approval.
Bingo.

This IS how it is done in some countries for L1 editors (they can only edit L1 segments, but those edits do not go live until approved.)

Sadly, one of the reasons that we are having MapRaids is because the local editors can not keep up with the influx of URs from this immensely popular app.
One of the reasons that ign_xxxx has come back in North America is because Waze KNEW that the existing editor team could not handle the addition of app-sourced Place Updates along with the normal workload.

Add to that the need to approve every edit done by every L1 or L2 editor and it will end up creating a situation where new editors will not be able to advance because their edits will never be approved.
taco909
Map Editor - Level 4
Map Editor - Level 4
Posts: 2246
Has thanked: 735 times
Been thanked: 638 times
Send a message
-- Rich

Post by taco909
irowiki wrote:With the trend in lock levels getting higher, low level editors can't edit anyway, so how do we solve that? Would be nice to have some sort of middle ground, even letting out of area high level editors approve edits, for instance.
The trend on lock levels getting higher is only for segments that L1s cannot edit anyways.
Places are a different story, and we could go back to lower lock levels on established places if Waze would eliminate this insane "trusted user" system.
taco909
Map Editor - Level 4
Map Editor - Level 4
Posts: 2246
Has thanked: 735 times
Been thanked: 638 times
Send a message
-- Rich

Post by vectorspace
Irowiki let me know about this post and his, so I thought I would reply in public...

I actually think it is a good idea in concept, but I think there are a couple of issues that would have to be overcome. If you can elaborate on how to overcome these issues, it might help those considering your idea for inclusion.

(1) Bemk92 already admitted that it could stack up a bunch of approval items. I don't know if this is good or bad. Irowiki noted that something a bit like this (but probably not as complicated) exists for edits to Places.

(2) This gets into a really big complexity of regression editing. Think about an editor that submits something like proposed in the idea, but it is not approved. Then someone else comes to edit that area. What are you going to do as the code developer? Are you going to lock that editing until that one lower level edit is approved? Or, are you somehow going to keep track of the next 10 subsequent edits to that area. If there are 10 more edits to that item after the first and a higher-rank editor comes in to approve the edits, which do they approve? How do you represent all those temporally stacked edits to help the higher-rank editor make a decision? With various editors coming in and changing things in that area, all with different levels, how do you handle that? What if a rank-6 editor comes in and edits the area without addressing the changes that are pending approval? How do you know how those edits effect the ones in the editing stack waiting approval?The point here is that some later edits could undo or modify, in complex ways, the edits for that area. If you're not careful, then chaos could ensue.

My guess is that few humans are smart enough to do this. In very limited environments, this can be solved. The WME doesn't seem to be very limited. It is sufficiently complex this may be an NP-complete problem (at least exponentially difficult problem).

A way to approach this is to limit the way such editing stacks work so they are simple and become simple to implement -- rather than to solve it for all possible situations. How would you suggest that be done?

(3) Is this really a priority issue, or just a cool idea? Waze has a whole queue of potential improvement and changes to make, so they cannot do them all. If this is really important, then you should think how to express that compared to other feature needs.
vectorspace
Posts: 1185
Has thanked: 99 times
Been thanked: 252 times
Send a message
vectorspace