bgodette wrote:I'm pretty sure you know my views on this by now and that that was for the benefit of the audience. But yes Waze does need a different security model, and given the direction they're going with tile processing and the new editor, it basically will have to become some sort of hybrid peer-review system. I'm still unsure about their intention to link editing rank to road importance (road volume) will work as they think it should as IMO it will overly restrict a new "more local" user for dealing with updated conditions on their daily commute vs the not-so-local AM/CM/Ranked.
I don't think a huge peer review process is the way to go. In the areas I drive there seem to be few active area managers (perhaps that has changed) and I expect most map changes are improvements or minor errors. One of my biggest bug bares is that I don't know how to get a list of "Problems in my Edit area" or even "Problems you have reported"; either of those would allow me to fix things far more easily without spending 5min scrolling around waiting to map I don't need to load! Anyway that is a separate rant. Here is the bits which relate to this topic.
I drove to work along a different route this week and as I was a passenger some of the time I was able to make more reports than usual. I would say I made about 10 reports of which half of them were correctly "fixed" by the editor. Of those the ones I managed to say Wrong Junction > "Roundabout" on top of what was shown as a T-Junction were most of the correctly fixed ones.
This shows two problems:
1) You can only convey a limited amount of information in the map error reports. This marker works to draw attention to a problem area but it does not always contain enough information for the editor to fix the problem. That is not the user or editors fault, you can only do so much while in a car.
2) The editors (I checked, it was more than one) in my area are very keen to edit and fix but not so keen to ask for more information. I expect this is in part down to Fixed = 1 point and More Info Needed = 0 points.
I think that two changes would help with the above. Firstly the email telling me my problem has been solved should also allow me to review the change and state if it is not. I think a reasonable amount of text should be required along with the reopen message to encourage the problem reporter to better explain what they wanted to see.
Now to stop the whole update system from grinding to a halt it is not necessary to prevent the update from going through, indeed it is probably an easier code change if it does not. A Simple by allowing a reporter to Reopen an edit, deducting 1 or 2 Fixed counts from the editor who erroneously marked it as fixed should mean the correct edit can then be made and the correct editor gets the only credit. By deducting 2 points for a bad fix it may also push more editors to check with the reporter first; 1 point off it nothing if they can just mark the same problem as fixed again and get that point back.