[Tagging] Access by permit

Kevin Kenny kevin.b.kenny+osm at gmail.com
Tue Sep 19 03:12:12 UTC 2017


On Mon, Sep 18, 2017 at 9:30 PM, Dave Swarthout <daveswarthout at gmail.com>
wrote:

> Unfortunately, it might never happen because even with what I see as
> crystal clear reasoning provided by Kevin, it's obvious to
> me consensus will not be possible within the group. But should it somehow
> come to pass that a better tagging scenario results, I can easily replace
> my access=permit tags.
>

Actually, I think we're getting pretty close; it appears that "map what's
on the ground" were magic words in this dialogue. What's on the ground is a
binch of posters saying "ACCESS BY PERMIT ONLY" and contact information -
and a line of survey blazes on the trees. Everything else: fees, seasons,
quotas, lotteries, whatever, is not observable in the field. I'll concede
that I've become quite cross at my inability to explain myself, but don't
let my temper convince you that consensus can't form.


> As for whether the object also contains contact info or permit info as
> well, that is an option available to mappers much like the number of lanes
> of a highway, or its surface - it's optional and can always be added later.
> To make such a subsidiary tag a de facto requirement for acceptance is over
> shooting the mark
>

I'm fine with saying that access=permit ordinarily SHOULD be paired with at
least some contact information for the permitting organization. It's rare
to know that a permit is required and not know who grants it. There's
always incomplete tagging, and data consumers always have to expect to deal
with objects that are underspecified, but let's set a baseline that gives
at least minimally useful information.

      - - -

Oh, and some meta-observations, only marginally on topic, so feel free to
stop reading.

When a data consumer says that they have an application that wishes to
render some objects one way, and some objects another way, at least that
consumer has some idea that the objects are different. Simply asserting
that the two are the same (that "permit" means the same thing as "private")
is telling that consumer, "you shouldn't want what your're requesting."
That is never helpful, even when it's correct.

When a data consumer comes here asking how to tag something that they wish
to render and use - and are willing to enter the data to support it -
that's a big gain for everyone. The consumer gets the problem solved, the
map gets the data, the community gains a member, and the problem is solved,
not only for that consumer but also for everyone else with the same issue.

Data consumers, the people who draw the maps, implement the routers and
produce the directories - or rather, their customers, the people who read
the maps, follow the directions, and look up the locations - are why we're
building this database, aren't they?

By contrast, the arguments about tagging among people that are mapping "on
spec" always strike me as faintly ridiculous. I simply stay away from the
discussions of how many angels can dance on a particular pinhead, when I
have no idea what application has a use for angel count or how it would
interpret that data to render a useful result.

There are only a few exceptions, I think. We've decided for a number of
reasons that cadastral data (except for administrative boundaries, public
and landmark land uses, and some other notable things) does not belong in
OSM. We've decided that historical data does not belong is OSM (because it
is not visible on the ground: visible traces may and should be mapped). And
people propose duplicate tagging schemes out of ignorance - so by all means
ask if an existing tag would serve the purpose. But please don't dismiss
people's requests out of hand. At least listen and try to understand what's
being said. Examine whether the thing might simply violate your cultural
assumptions, for instance. From highway classification to administrative
boundaries to default property access, another country's practices might
simply be different from yours.

That's the source of a lot of our trans-Atlantic friction, by the way. US
practices are best described, in almost all areas, as 'untidy'. The people
who do wonderful work in neatening the categories and descriptions into
which we sort our geographical objects are appalled at this, and rightly
fear that the map will turn to chaos. In turn, many USAian mappers hear the
people who call desperately for order as saying 'our data model is fine;
fix your country!' (Would that it were that simple!) Yes, the US is an
messy, even chaotic, country. Many of us like it that way. In any case, we
have to model it as best we can. If that means using the 'national park'
tag on something (http://www.openstreetmap.org/relation/6265477) that is
surely not a National Park, nor even national, and is arguably also not a
park, so be it. Nothing else fits, and the underlying object is both
one-of-a-kind and locally extremely important. If it means describing a
type of land access restriction that is common here but (apparently) rare
elsewhere, let us have it. If it means that we have county boundaries that
are mapped even though they don't exist on the ground in any way, well, we
do have indefinite boundaries even at the county level. (Example:
http://caltopo.com/map.html#ll=43.62167,-74.51014&z=15&b=t The boundary is
back in the woods, in an area that's never been settled, and nowadays is
all protected wilderness. If it ever becomes important, a survey team
and/or a court of law can settle where it is. In the meantime, the locals
see 'indefinite boundary' as not being worth the money to fix.) Some US
tagging will always be founded on little lies - because the truth is
inconsistent or indeterminate.

We're all trying to map the world. It keeps turning out to be a bigger and
stranger place than any of us imagine.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20170918/ae7fb8ac/attachment-0001.html>


More information about the Tagging mailing list