I can think of two scenarios where it might be sensible not to permit any one to edit a particular tag. The first one: Currently a highway can have a speed tag associated with it. Technically a car can have its speed limited by its GPS position.
One issue they have is how to indicate the speed limits. A map perhaps with a speed tag on the highway? Technically we have this available in OSM.
City planners get excited about the idea of traffic calming without physical measures. Insurance companies can see benefits especially for younger drivers. However a technically savvy 16 year old could very easily adjust the speed limit in OSM around their school if OSM was used as the basis for speed control so they could speed to school. Speeding school kids are a problem on my street, we have two high schools one at each end, a 40 km/h speed limit but 90 km/h when the schools finish for the day is not unknown. Therefore some restriction on who could amend the limit might be useful.
Or an alternative might be a table of speed limits and road names, you take the position from OSM which gives you the road name but you may have different speed limits on different sections.
The other one is automated_external_defibrillator yes tag it but it can be life threatening if some one maliciously changes its location. If you need one of these then you need it in a hurry and any delay can be fatal. Having been involved in computer security for quite a few years it would be naive to think that some one won’t enjoy a feeling of power if they maliciously modified the location of one of these.
Whilst OpenStreetMap is certainly becoming an amazingly powerful dataset I think restricting editing of tags would be wrong. We have measures to detect malicious edits and revert them although the deliberate and one-off vandalism you’re talking about would be very hard to detect.
OpenStreetMap is perfect for many applications but there is no guarantee of accuracy and so it shouldn’t be used in situations that are incredibly sensitive like the ones you are suggesting it might be used for.
What you could do is maintain a mirror of OSM using minutely diffs but put all modifications to maxspeed and automated_external_defibrillator keys on hold until approved by administrators.
This improved OSM database could then be offered as a service for the innovative uses you describe above. However the process of approving changes to specific tags does not belong in OSM itself.
OSM is primarily a database for collaboratively collected data, not for official or authoritative data. In my opinion, its primary virtues are easy and frequent updates and the absence of hard limits for the depth and breadth of information added to the map (and, of course, its free license). OSM doesn’t provide guarantees for correctness, and it doesn’t provide uniform data quality.
As such, it currently isn’t suited well for the applications you describe. It might be possible to make OSM data usable for these purposes, but any attempts to do so must not negatively affect our current strengths, which are, imo, the foundations of our crowdsourcing approach. Preventing mappers from editing certain tags would certainly create additional barriers for beginners, as well as potentially affect the speed of updates and corrections. It would also require a privileged group of contributors and might be an additional source for conflict within the project.
There might be options for quality control that don’t have these disadvantages, though. I could imagine something like MediaWiki’s flagged revisions being used - versions of objects could be marked as “approved” by members of a certain group. For example, I could create a group “The Leet Maxspeed watchers”, and everyone who trusted my group’s quality criteria could use only those highways (or versions of highways) that were flagged as
correct by a member of my group.
To be compatible with OSM’s avoidance of central authority, however, it should be possible for someone else to found the “Ultimate Maxspeed QA”
with its own approval flags. Furthermore, it should be possible for someone who doesn’t need that level of quality to simply get the latest version of
the data without caring for quality flags at all. These two aspects would be clearly different from MediaWiki’s implementation and avoid most of the conflicts the introduction of flagged revisions have caused for example in German Wikipedia.
Of course, it wouldn’t be trivial to implement all these features.
Editing should in my view not be restricted, as this will go to an endless list.
In Germany they are working on maps for disabled. This means that traffic lights, stairs, telephone cells, ATM’s etc become important items for that map. Just as with speed limits, a wrongly placed traffic light can have consequences.
There is also some experience with individuals who tried to get their way by modifying the map in a non-agreed way and that has been resolved as well.
Reasons enough not to limit the freedom to modify tags and roads.
The type of vandals and people that get a feeling of power (or whatever kind of satisfaction they get) from vandalizing or doing malicious deeds don’t really get stopped by things like what you are suggesting.
It may have the opposite effect, where these people would view it as a challenge and be more likely to try.
As an example: name one online game that does not have cheats even though there are many anti cheat systems in place (punkbuster, steam, etc.)
Really, they should instead realize how much “power” they are being trusted with up front.
And if they abuse that, hopefully someday they will realize that they are ultimately doing themselves the most wrong.
If nothing else, hopefully Karma will bite them in the butt.
as a side note, If I was one of those kids with a gps limiting the speed of my car, the first thing I would be doing is “modifying” the gps or the car.
