[Talk-us] Chesapeake Bay Bridge reversible lanes

[Paul Johnson]

On Thu, Mar 23, 2023 at 12:46 AM Alan Brown via Talk-us <
talk-us at openstreetmap.org> wrote:

> Hi,
>
> Haven't posted to this for nearly ten years, but I figured this may be the
> best place to get an answer to my question.
>
> It's a question here about handling the Chesapeake Bay Bridge reversible
> lanes.
>
> The Chesapeake Bay Bridge has two major part, one represented by
> https://www.openstreetmap.org/way/27164664, another by
> https://www.openstreetmap.org/way/27164663.
>
> Most of the lanes most of the time on the former side go in one direction,
> and  the lanes on the latter side go the other direction.
>
> However - if we look at the tags on
> https://www.openstreetmap.org/way/27164664, we see the following:
>
> oneway <https://wiki.openstreetmap.org/wiki/Key:oneway?uselang=en-GB> yes
> <https://wiki.openstreetmap.org/wiki/Tag:oneway=yes?uselang=en-GB>
> lanes <https://wiki.openstreetmap.org/wiki/Key:lanes?uselang=en-GB> 3
> lanes:both_ways
> <https://wiki.openstreetmap.org/wiki/Key:lanes:both%20ways?uselang=en-GB>
> 1
> lanes:forward
> <https://wiki.openstreetmap.org/wiki/Key:lanes:forward?uselang=en-GB> 2
>
> So one lane is reversible.
>
> We see this here:  https://goo.gl/maps/72nEj58Qi8xiUsad6
>
> The lane is separated from others with a double yellow line - so drivers,
> regardless of which direction they are going, should not enter or exit it.
>
> Should this lane be digitized separately, tagged as reversible?  Even
> though it's the same bit of pavement, it would allow routing engines that
> do not have lane specific routing to handle this situation.  Or there
> another way to model this?  It seems to me that the lanes:both_way tag in
> combination with a oneway: yes tag would be very confusing to interpret;
> they seem to be at odds with each other.
>

No.  If I were tagging that, I'd change oneway=yes to oneway=reversible
<https://wiki.openstreetmap.org/wiki/Tag:oneway%3Dreversible> or a specific
conditional value if the time things reverse changes.


> While we're at it:
>
> I know in the past highway: motorway and highway: motorway_link ways were
> to be interpreted as one-way in the direction of digitization if there was
> no oneway tag-value pair.  The latest documentation recommends marking
> two-way motorway_links with oneway=no.  However - currently - in the
> absence of any oneway tag, should we still assume it is oneway?  In the
> past, it was dangerous to, as the data was not fully attributed.   But we
> have found some two-way motorway_links that do not have oneway=no.  We can
> correct and tag these, but in the meantime, what should we do?
>

When in doubt, explicitly tag what's going on.  Don't guess what consumers
are going to assume, tell the consumers what reality is.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20230414/6e5d2f8b/attachment.htm>