To: talk@openstreetmap.org From: ajt1...@gmail.com Date: Thu, 28 Jan 2016 19:50:11 +0000 Subject: Re: [OSM-talk] Applicability of wiki tagging and votes: may, should or must More seriously, any dataset that has no rules enforced at the API level must be assumed to have data in it that doesn't meet a specification that is written down somewhere, but not enforced. Someone wrote that wiki page long ago but didn't actually do anything else, presumably expecting the magic code and project management fairies to look after all the other changes that they expected to happen. I know that the free tagging scheme is one of the main strenghts of OSM, and I'm aware of its advantages for modelling the real world, which per se contains elements that will never fit in a strict tagging scheme. My point is, there are multiple references (editors, Wiki, MLs, and, maybe a bit consumers), which regularly contradict one another about what they assume to be the "correct" tagging; it would be far more coherent to have a single reference, that users SHOULD follow whenever possible, which would be the start of each debate, transcript its end and the decisi,ons made, and which would centralize tagging defs and their modifications. This way, the undocumented tagging could be reduced, at least the unnecessary, unnecessarily confusing part, and increase the DB usability for consumers, who would be less required to adjust to the different, virtually unlimited, existing tagging schemes. Note: I use SHOULD as defined in the IETF RFCs (https://www.ietf.org/rfc/rfc2119.txt), that is: mean that there may exist valid reasons in particular circumstances to ignore a particular item, but the full implications must be understood and carefully weighed before choosing a different course.
_______________________________________________ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
