New taginfo features and a challenge

I just added a few features to taginfo: There is a new “Similar” tab on all key pages, see for instance all keys similar to the “highway” key. This makes it easy to find related keys and typos. The new Similar keys report shows keys that are commonly used together with similar keys that are rarely used. Often, but not always, the rare keys are just misspellings.

Another new feature is a statistic that shows how many unique keys, tags, and relation types there were over time. As you can see in two years we went from an already high number of about 40.000 keys to over 55.000 keys. Looking at those keys it is obvious that many of them are just typos or other errors. Sometimes it feels rather difficult to find the useful keys in that sea of bogus ones. But now with the new “Similar” tab and report and with the long-existing Characters in keys and Key lengths reports it should be easy to find lots of candidates for cleanup. And the “JOSM” and “Level0 Editor” links on the key pages make it easy to jump into an editor and fix those cases.

So lets clean up the OSM database! I challenge the OSM community to get back to the 40.000 keys from two years ago and do it before the end of March. Fix those typos and clean up obviously wrong keys. But remember to be conservative: Just because you don’t understand the meaning of a key or because it is not documented, it doesn’t have to be wrong.

UPDATE: Because the problems with mechanical edits are already starting I have to add this: Do NOT just use search-and-replace to fix those typos. Please look at each and every object you want to fix and make sure you leave it in better shape than it was before. If you are unsure, do not change the object. Objects that have one tagging problem often have several, if you only fix the one typo, you make the other problems harder to find and you make the object look newer to other mappers (“Oh, this object has been recently edited, so it is probably correct.”). Also think about it this way: Each time you edit an object, this object will have your name attached to it. And even though you might have only changed one aspect of the object it will look like you endorse the whole object (“User X has looked at this object and edited it, so he must think it is okay.”). Other wrong tags on the object will look more valid, because there are now more users (you) that seemingly used those tags. So look at each case separately, it will be a bit more work, but it will improve the OSM data immeasurably.