[Tagging] Rules (was: Feature proposal - Approved - deprecate embassy=embassy)

[Kevin Kenny]

On Thu, Feb 3, 2022 at 9:08 PM stevea <steveaOSM at softworkers.com> wrote:

> I ask each of us to manage as best we can the precious resource OSM has to
> "grow well."  This means understanding, in an adult, realistic way, that
> "we emerge."  Tagging in OSM is sometimes "it's done like this a bit more
> here" (and there).  "Managed" globally, like everything can always be
> watched?  Ladies and gentlemen, this project is organic.  It will have
> mushrooms and what some call weeds and others call noxious plants.  Let's
> get busy managing that growing this project requires that we not only
> understand that, but live it.  Harvesting well requires we sow good seeds
> and till the land with generations of future ahead of us.  We cannot fail
> to innovate, we cannot endlessly stifle.  Besides, mushrooms play a vital
> role, and some are delicious.
>

To continue Steve's flowery metaphor, when a plant grows where a gardener
didn't place it, it's called a 'weed' if the gardener doesn't like it, and
a 'volunteer' if the gardener does. OSM is created, maintained, grown,
curated, administered almost entirely by volunteers.

Also, "he who does the work makes the rules."

Most tagging discussions I see on the mailing list rapidly devolve into two
categories when I have my data consumer's hat on. "I can ignore this
thread, because I can't imagine a rendering, routing, navigational or
planning application that I'd produce that would depend on this
distinction", or "there's a new synonym arising, and I'm going to have to
start looking for _this_ tag in addition to _that_ one."  In both cases I
shrug and move on.

The minority of tagging discussions that I get passionate about are the
ones where they will affect how I render or interpret something. Those
often end in a failed consensus. Generally, the proposal founders on
several things: (1) People won't accept the initial tagging but demand a
more detailed ontology, often one that depends on information that I can't
readily observe in the field; (2) People don't understand the difference
between the objects that I want to distinguish and simply tell me "you
shouldn't want that." (3) People from elsewhere in the world who simply
cannot believe that there is an object with the characteristics that I want
to map.  There's also a category of comments that just mystify me: "I want
to render a map that distinguishes objects of class 1 and objects of class
2; what tagging ought I use in OSM to describe the distinction?" being
answered by, "Don't tag for the renderer!"

Concrete example : I don't really care when rendering a map whether the
pair of ruts that I spotted in the woods appears in the database as
'residential', 'unclassified', 'service' or 'track'.  If the way is in the
database and has physical characteristics, I can render "a minor road
without a hard surface" or route accordingly. But I definitely bristle at
being told that I can't map such a way as anything other than
`highway=road` unless I know where it ultimately leads. Getting the tag
wrong among those classes in that situation will have nearly no impact on
any routing, navigation or rendering.

Only one or two similar discussions have ended in enough acrimony that
there are one or two classes of object that I will not map. I can live
without them for now.

Occasionally, I'll jump in and advise on decisions that I think are
ill-considered. (For instance, I don't think that deducing anything
affirmative from the absence of a tag is ever going to end well.)

But for the most part, 'any tag you like' works fine, at least for the data
that I consume.  If I don't understand tagging, that's either a fault in my
data-consuming code (I occasionally troll about for common tag patterns
that I'm missing), or else something I can safely ignore. There are a lot
of mappers who are busily adding things that I don't find informative. I
can ignore them.  There are map data added with tagging that I find
misleading. I can work with the mappers. In the end, I get a map that I can
use, and that's what matters.

We have a messy map at least in part because we inhabit a messy world.


PS: For actual "how do I tag this thing?" questions, I've learnt through
experience that taginfo is more informative than either this list or the
Wiki.  I've gone astray in the past by taking the Wiki's advice. And on the
tagging list, where there are two mappers, there are three opinions. Our
arguments grow so heated precisely because so little is at stake.

-- 
73 de ke9tv/2, Kevin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20220204/8edb5914/attachment.htm>