Reading this thread from the Tagging mail list, I’ve noticed several posts describing a problem of lacking governance, leading to endless (even looped) discussions and other negative consequences for the project. I agree with that because if we really want to create, maintain and improve “the best map of the world”, it is counter-effective to rely on natural evolution only. Obviously, it will take too much time.

But I don’t think that governance requires a government as a group of people, like Mark Bradley have proposed in that thread. It could be enough to have a formal guideline or a declaration of goals. For example, we can endlessly argue about semicolon-delimited values, presenting nicely polished pros and cons. But it is impossible to have a consensus if we don’t have a goal or a guideline to test a certain proposal or a statement against it.

Like, how important is it to be able to query any new tag with an Overpass API? Is Overpass API a key part of the OSM infrastructure, or not? Should we care about having a universal meaning of every tag in any country? Should we care about not having a non-verifiable, vague or relative definition of a tag? What about overlapping definitions? Is it important not to force a data consumer to conduct a detailed research on every tag before he’d be able to use it? Is it acceptable to have a scheme that requires a pre-processing with a spatial query?

Without a common goal, any discussion inevitably degrades into a contest of personal views. A special group of people can’t fix it because they also need something to test everything against it. They can, probably, try to replace it with their wisdom, but it doesn’t seem to be a good idea.

Guidelines could be developed as a GitHub-hosted document(s) to be able to actually develop it and to maintain full control over it, instead of using a bit anarchist Wiki style.

There is an idea (also mentioned in that thread), that everything in this project is already governed by an implementation. But actually, it says nothing about how good or bad a certain thing is. For example, there could be an awful tagging scheme, but there could also be a person, who doesn’t think so, and who have implemented it in his tool, renderer or editor. Other people, who have not enough knowledge to understand what’s wrong with this scheme, just starting to use it (tag objects with it). Finally, we have a lot of objects tagged this way and it is nearly impossible to change this scheme because nobody cares except ones who keep saying that “changing things is the worst thing”. Doesn’t seem like a productive practice because it literally translates into a “the first one is the winner” principle - obviously, it makes any further improvements significantly harder.

As a bottom line, I want to say it again: government as a group of people doesn’t seem to be necessary but having better guidelines does.

---

Update. The difference between a government group and a guidelines development group is quite simple and obvious.

The government is an executive organization. Usually, it looks into every case within its scope and has to make a decision how to deal with this specific case. Obviously, it takes a lot of time and it is still impossible without having a commonly accepted set of rules. Otherwise, it leans towards an authoritarian system.

Hypothetically, guidelines development group is a “legislative” organization. It has to make decisions regarding of technical questions before such question will raise. The solution gets a power of rule. So, it solves a lot of similar future questions at once. It saves time, it can’t be personal, it is evidence-based.