Disappointment about WAZE

[krankyd]

Hello again llayden.
Even though I appear as 'waze champs', I'm also a part of the waze team.

I think as within many other forum posts, the debate slides to many different directions, as there are many 'open items' on waze. I'll do my best to keep mine focused. Before I do, I just wanted to mention that we recently had 2 big meetings with users (one in Europe and the other in the US). The purpose of these meetings was to understand the most important needs of users, and get a list of items we can follow up on and commit to.
You can check out the related threads here and here.

There are many important items that are not on the list, because as I said - there are still many things to add / improve / fix. There will always be such items. We do our best to keep the priorities balanced between the community and the company needs.

As for what you wrote:
1. Working on excessive nodes is indeed painful, but that's not the most important things to get your maps closer to perfection (an additional node has its affects, but routing is not one of them. There's a minor penalty for crossing a node, but when there are no other options in the node, it's null). We think it's much more important to improve the process of creating new maps, and simplifying the editor, and these are two 'heavy' tasks we've been working on.
2. I don't know about adding too many options to the paving panel. After all you are driving while paving, and safety is a measure issue for us. Naming the roads is currently disabled on version 3.0, but when it comes back, I think your suggestion of naming the road as soon as you're done driving is a very good one. I'll forward it to the team.
3. The statistics by country have been discussed in both meetups I've mentioned before, and we are currently working on something that will allow just that.
4. Following the European meetup, we've implemented a process that disables the option to add new cities. Currently only enabled in the Netherlands, but we plan on enabling it world wide (giving the option to add new cities for users with editing rank of 5 cones and above). The new editor does not allow adding a segment to a city which is too far away from the segment.
5. We had this operation with the cartouche (the old editor) - at least partially. It created some overhead for us and for editors. What we're planning for the new permissions system is enabling segments for editing based on your own rank. If there's a highway which is traveled by more than X wazers per day, only a wazer with editing rank X will be able to edit. We assume once you can edit, you are also allowed to edit.
I know AMs today get access to unlimited editing with a very easy finger on the trigger. Again, part of the new permissions system will change that.
6. We're considering making editing rank 6 cones 'global editors' as you suggested. I personally think this is a good idea.
7. Not sure how easy this will be (implementing a layer of 'wazerability'). But I'll pass this on and see what others think.
8. I think the new format of UR (showing you the drive + suggested route) come close to what you're saying. From our experience adding too many routing options does not work well. We are working on removing the 'prefer cookie munching' option (which, by the way, does what you requested - other than the part about the destination).

Thank you

[gettingthere]

Regarding granting global editing rights:

Although most editors that achieve 6 cones are highly experienced editors and will likely not just start editing blindly in others countries, I think that Waze needs to be cautious with this approach.

There are many barriers to editing in other countries:

* driving direction (driving on the left vs the right)
* language
* local editing standards (which may be documented in the forum and/or wiki in another language)
* road construction standards
* typical turns allowed/denied per certain road layout
* finding data that is not restricted for getting road info, street names, etc ( again, possible language barrier)
* etc

We all know how the IGN editors fared in editing in countries that they were not familiar with.

I was a very vocal supporter in getting editors with a level 4 editing rank country manager editing access. In addition, I am all for more editors working the map.

With this said, I suggest that Waze proceed with caution before opening up multi-country editing based on editor rank or number of cones.

[llayden]

1. Working on excessive nodes is indeed painful, but that's not the most important things to get your maps closer to perfection (an additional node has its affects, but routing is not one of them. There's a minor penalty for crossing a node, but when there are no other options in the node, it's null). We think it's much more important to improve the process of creating new maps, and simplifying the editor, and these are two 'heavy' tasks we've been working on.

I think you've missed the point here. Here's the example that drove the point home to me. I paved a road 4.6km long. When it appeared in my archive a day later, it was made up of 58 segments. Waze usually does better than that but that's what I got on that occasion - an average of less than 80m per segment with some as short as 12m. They were all one way segments. To get this short length of road working I had to do the following (using Cartouche): I had to ctrl/click 58 times (the select all roads in between option did not work - I only ever got it to work twice). I then changed them to two way, named them and set the road type. To ctrl/click 58 times had its complications too. If I attempted it zoomed out, I kept getting junctions. Just why Waze will let you select roads and junctions when there is nothing one can do with both selected, I don't know but it does (un)helpfully tell you that you have done so! So I had to zoom in far enough to avoid picking up junctions. So now I click and drag the map left - repeat 58 times. However, having gone through these exertions, I am not even nearly finished. I am left with 58 segments of named bi-directional roads separated by 57 uni-directional junctions - I had show connectivity switched on (which many novice editors do not) and which Waze again (un)helpfully defaults to off to punish you and the users for you not reading the Wiki properly. Hmmm! So after another 57 clicks and occasional drags to the right, I go to the Enable all turns option to find it greyed out. Those nice Waze people, to make me grateful for being able to do it on segments have prevented me from doing it on junctions. (I'm sorry I cannot resist sarcasm). So now I start doing them one by one. The process is: Click a junction - move mouse half way across screen to enable all turns - move mouse back - click and drag map slightly to left - repeat 57 times. Now I connected the two road ends - job done, or so I thought. Of course, a month later I discovered I couldn't route backwards through that road - I had missed one or more (of the 12m apart ones). So another 57 clicks . . . Then the thought crosses my mind that loads of roads have been done in this way and of course I look and find hundreds of them - in fact far more wrong than right. So I eventually consider two strategies - remove the junctions or bridge the roads. Removing the junctions alters the geometry at the point removed so bridging becomes my preferred method and I do this a lot - every time I find a uni-directional junction on a two-way road I remove ALL the artificial junctions. Its only slight slower than enabling all junctions and saves me and other suckers from having to work all the way along the road checking the junctions yet again when another route fails. Personally I have paved 1533km. Assume my example above is three times worse than average - say average lengths of 250m. That's about 12,000 edits, just making one-way into two-ways and enabling all turns. Of course, I don't just edit my own roads - so that 12,000 becomes multiplied by 3..4..5..6 ?? Your guess is as good as mine. In fact my 40,000 estimate now looks low, and every one of them was necessary - not cosmetic as you suggest.

If there was a single favour you could do for Irish editors, it would be to take every junction in the country and enable all turns on them. The work in putting back real turn restrictions would be tiny in comparison to the work finding and enabling the incorrect ones.

Even for manually added junctions, the story is not much better. Unless you tick the enable all turns box before adding the turn, you have a problem. It does stay ticked for future junctions - until you hit refresh on the browser because everything has frozen. However, an unaware editor is adding hidden time bombs to the map? Why do all the defaults seem to be chosen to maximise the work of editors? (rhetorical question - please don't feel the need to answer)

2. I don't know about adding too many options to the paving panel. After all you are driving while paving, and safety is a measure issue for us. Naming the roads is currently disabled on version 3.0, but when it comes back, I think your suggestion of naming the road as soon as you're done driving is a very good one. I'll forward it to the team.

You must know that your answer here is weak. The client allows you to record, for example, traffic jams - click button bottom right, click button on panel that displays, choose between moderate/heavy/standstill, optionally take a photograph and write text to back up your info, click send button. All this activity is encouraged by Waze and rewarded with 6 points. But four buttons across the top of the app while paving a road is dangerous? Hmmmmm.

3. The statistics by country have been discussed in both meetups I've mentioned before, and we are currently working on something that will allow just that.

good!

4. Following the European meetup, we've implemented a process that disables the option to add new cities. Currently only enabled in the Netherlands, but we plan on enabling it world wide (giving the option to add new cities for users with editing rank of 5 cones and above). The new editor does not allow adding a segment to a city which is too far away from the segment.

Maybe I can help. I'm the CIO of a group of freight companies and somewhere within our database is a list of all the cities/towns in the world with a population over 2,000 with their Latitude & Longitude and population. We got it from open sources. You can have it if you want to seed a database table. I would suggest you listen to gordonski and use the opportunity to cater for different levels of organisation in different countries (Quote: "In case this will be done, please do not forget countries where streets don't only belong to a city and a state, but e.g. to a town, a community, a district town, a state and finally the country. I know, we Germans have a tendency to over-organize things, but we need to deal with it some way and could need one or two extra layers here...")

5. We had this operation with the cartouche (the old editor) - at least partially. It created some overhead for us and for editors. What we're planning for the new permissions system is enabling segments for editing based on your own rank. If there's a highway which is traveled by more than X wazers per day, only a wazer with editing rank X will be able to edit. We assume once you can edit, you are also allowed to edit.
I know AMs today get access to unlimited editing with a very easy finger on the trigger. Again, part of the new permissions system will change that.

I think the only problem here is when digital photos are out of date. No experienced editor is going to make a mess, except by accident - so some indication/note on the map would cure that problem.

6. We're considering making editing rank 6 cones 'global editors' as you suggested. I personally think this is a good idea.

That's not what I said. My threshold was 20,000 points and would only apply in the absence of an AM. I wouldn't let people edit outside their country, without applying explicitly for permission to do so.

7. Not sure how easy this will be (implementing a layer of 'wazerability'). But I'll pass this on and see what others think.

Strictly speaking this is off topic. What I was trying to do here was give end users some sense of hope!

8. I think the new format of UR (showing you the drive + suggested route) come close to what you're saying. From our experience adding too many routing options does not work well. We are working on removing the 'prefer cookie munching' option (which, by the way, does what you requested - other than the part about the destination).

The problem here is that users don't log the problem until after the event and, given safety considerations, than can be up to a kilometer away. Papyrus typically shows me intended path and actual path the same. In Cartouche, I was on top of the situation for Dublin every day. So a user logged a problem. I either sorted it or closed it explaining to them that I couldn't pinpoint it. Frequently, they would turn on Waze and log the problem again (at home) saying for example no right turn from x onto y (problem now logged miles from where it actually happened). I would then fix it, close it and thank them. Dialogue was actually happening. Now they get no feedback and I can't get any more info. (Desperately suppressing sarcastic comment).

My thought here was also off topic - give the Wazer a sense of adventure when Waze fails to route them.

To help the map though, having an automatic UR when a user goes through a turn restriction (whether being routed or just driving) and/or a warning on the client would be very helpful in finding the hidden restrictions.

I'm sorry that this was such a long winded post. I do have a lot more to say on these points. But perhaps my points are clearer now.

I must now go and buy a cake and a candle - I'm a one year old Wazer today.

[dmcconachie]

Don't have time to respond to your whole post but ... stop using the old editor! Wouldn't solve the problem as you describe but it's far better! Part of the reason Waze are pushing the new editor through, albeit with a few limitations, is that the old one is causing database problems!

And also regarding the same, why go through and select each inidividual segment first? Would make more sense to delete each node first then deal with one segment! Again, not a perfect solution but a step quicker!

[RallyChris]

The one way problem, is a fail safe solution.

This is following the seen it done this way, so until someone tells me differently, or I see it the other way, this is the way it is method.

Yes, it may be time consuming for an editor to correct things up, but what if the editor is not as diligent, and is really paving a one way road. The update process can fix turning a one way into a two way. It doesn't (or at least not suppose to) turn two ways back into one.

[gettingthere]

A great example why paving roads should only be used as a last resort. Only pave when there is extremely poor or no imagery and no gps points to use to draw the roads directly in the Waze Map Editor.

[bgodette]

A great example why paving roads should only be used as a last resort. Only pave when there is extremely poor or no imagery and no gps points to use to draw the roads directly in the Waze Map Editor.

Don't even pave at all, just drive and create the missing segments from your Drives in the new editor. No excessive junction nonsense.

[llayden]

A great example why paving roads should only be used as a last resort. Only pave when there is extremely poor or no imagery and no gps points to use to draw the roads directly in the Waze Map Editor.

Exactly the situation I was in - and doesn't address the fact that heaps of other people had done so already.

[llayden]

Don't have time to respond to your whole post but ... stop using the old editor! Wouldn't solve the problem as you describe but it's far better! Part of the reason Waze are pushing the new editor through, albeit with a few limitations, is that the old one is causing database problems!

And also regarding the same, why go through and select each inidividual segment first? Would make more sense to delete each node first then deal with one segment! Again, not a perfect solution but a step quicker!

I have highlighted the important bit of your answer! The example was from last July. I don't pave any more - I create from the archive or I pave and delete and then create from the Archive so the road is in the client but the problem remains with so many people paving and so many non-return valves all over the place both from paving and poor editing.

... and BTW I do still use the old editor because I don't like the new one.

[llayden]

A great example why paving roads should only be used as a last resort. Only pave when there is extremely poor or no imagery and no gps points to use to draw the roads directly in the Waze Map Editor.

Don't even pave at all, just drive and create the missing segments from your Drives in the new editor. No excessive junction nonsense.

Yes. But how about all the roads paved by people who never edit?