[Tagging] Feature Proposal - RFC (etc) for crossing:signals

[Nick Bolten]

> It is?  I hadn't noticed.

Yes.

> I take a very different view, that crossing=traffic_signals says that the
crossing is controlled by traffic signals.  There may or may not be
markings.  Those markings may or may not be similar to markings at
crossings without traffic signals but, if the lights are functioning those
markings have no legal significance and do not determine rights of way.

This is still not true. Cars cannot occupy that space, per law, in my area.
I assume this is probably true in yours as well.

I'd be happy if everyone had this interpretation, however, because it would
make my job of arguing about orthogonality very easy.

> I, like some others here, think it rather obsessive of you to insist on
mapping what we consider to be an irrelevancy.

First, we already map this. It's what the tag crossing=uncontrolled means,
per the wiki: a marked crossing, like a crosswalk in the US. You have to
put aside this idea of what "controlled" and "uncontrolled" mean, they are
apparently irrelevant. Perhaps we should change the schema so this kind of
error stops happening...

Anyways, that's a strange way to frame "mapping something I don't care
about". How is it obsessive? I've already listed several important use
cases, so I will be blunt: do you think people with low vision are
irrelevant and don't matter? Is this an ableist community? Do pedestrians
getting struck by cars not matter? Is it okay that they die?

> The crossing is CONTROLLED by the lights and that is the important factor.

This is completely unstated in the tag definitions. It's not actually the
important factor, per the most official-ish sources we have. Clearly
absolutely everyone (including you) is confused about how to use these tags.

> Sure, if you can come up with something that isn't disruptive and has
other benefits, then it MAY be worth coming up with a tagging scheme that
allows
us to indicate whether a crossing controlled by lights also has markings.

I did. It's this proposal. You should check it out, it makes a
multi-pronged argument regarding the problems of this tag and benefits of
changing it. I'm sure the proposal could always be improved, and I'm very
interested in feedback, but continually rehashing inaccurate and myopic
misrepresentations isn't productive.

> At best, all I've seen indicates that maybe editors should make it
clearer to mappers if they change a crossing tagged as traffic signals to
one with markings that perhaps they're
using aerial imagery to undo what somebody has verified on the ground.

There are currently four headings pointing out problems with
crossing=traffic_signals, several follow-up arguments about disambiguating
the whole traffic_signals namespace, definitions of what the new schema
would be, and a simplification of mapping down to two straightforward
questions (extremely easy to support in an editor) vs. the current
confusing trade-off schema.

The example you list is worse than you've stated, though. The *original*
mapping can easily be from aerial imagery and state crossing=uncontrolled.
Because the crossing subtag is now populated, QA tools / Overpass Turbo
will no longer pick it up as unset and needing data. Only a thorough
in-person audit / broken use case will detect the error.

> I don't deny that such edits may be a problem, I'm not convinced your
proposal is the best solution.

I'm open to other strategies. What do you propose?


On Mon, May 20, 2019 at 5:53 AM Paul Allen <pla16021 at gmail.com> wrote:

> On Mon, 20 May 2019 at 06:53, Nick Bolten <nbolten at gmail.com> wrote:
>
>>
>> This is topical, as crossing=traffic_signals is often claimed to imply
>> crossing=marked.
>>
>
> It is?  I hadn't noticed.  I take a very different view, that
> crossing=traffic_signals says that
> the crossing is controlled by traffic signals.  There may or may not be
> markings.  Those
> markings may or may not be similar to markings at crossings without
> traffic signals but,
> if the lights are functioning those markings have no legal significance
> and do not
> determine rights of way.
>
> I, like some others here, think it rather obsessive of you to insist on
> mapping what we
> consider to be an irrelevancy.  The crossing is CONTROLLED by the lights
> and that
> is the important factor.  Sure, if you can come up with something that
> isn't disruptive and
> has other benefits, then it MAY be worth coming up with a tagging scheme
> that allows
> us to indicate whether a crossing controlled by lights also has markings.
>
> At best, all I've seen indicates that maybe editors should make it clearer
> to mappers if
> they change a crossing tagged as traffic signals to one with markings that
> perhaps they're
> using aerial imagery to undo what somebody has verified on the ground.  I
> don't deny that
> such edits may be a problem, I'm not convinced your proposal is the best
> solution.
>
> --
> Paul
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20190520/3eb074da/attachment.html>