[Tagging] noexit=yes : the outcome

fly lowflight66 at googlemail.com
Mon Apr 14 12:54:09 UTC 2014


Am 14.04.2014 14:07, schrieb André Pirard:
> Hi,
> 
> Side note: Please note that I just found two versions of the "Use the
> noexit=yes tag..." text. One ahead that says that it must be used only
> on nodes and the former one inside that says that it can be used on ways
> too.  I left the newcomer where it was: ahead but stroked out at your
> disposal.  Fun, fun, fun.
> 
> So, Pieren added to that page that noexit=yes can be tagged on ways.
> He did this because contributors do that tagging, "open-mindedly".
> Friends pointed out quite rightly that what the specification explains
> cannot possibly be tagged on a way.  It's inescapable.
> Insistence to make the impossible come true generated an endless,
> fruitless discussion.
> 
> As I find that understanding the meaning of a tag on the map is
> important, I pointed out that no one had asked why the contributors
> tagged noexit=yes that way and that the only clues came from those who
> spoke up: "I had not consulted the specification [for a long time]", "A
> friend recommended that", 'I thought that I was tagging a No Exit sign", ...
> Except for the last, they don't explain the meaning of their tags, but
> one thing appears certain: the meaning is different from the
> specification which (btw/because it) is vastly misunderstood.
> And what else, if they don't have the subtleties of the specification in
> mind, could it be that "the No Exit sign"?
> 
> So, to make this explanation short and kill two birds with one stone,
> please read <http://wiki.openstreetmap.org/wiki/Key:noexit>, I modified
> the wiki as follows:
>> The noexit=yes tag has two meanings according to whether it's on a
>> node or on a way:
>>
>> On a node:  present text, but  "Use ... on a node"
>>
>> On a way:  Use the *noexit*=yes tag on a way Way
>> <http://wiki.openstreetmap.org/wiki/Elements#Way> to indicate that the
>> way is leading to a dead end.
> 
> Remarks (here):
> 
>   * Please note that it would logically be necessary to indicate which
>     direction leads to the dead end.
>   * The second definition is not exactly complete. Some contributors tag
>     noexit=yes all over a small, meshed neighborhood that contains a
>     dead end road, but not necessarily if there's only one entry road.
>     Well, putting it short again, a more appropriate description in that
>     case (when the said way/road crosses junctions) would be: "to
>     indicate that one will have to return to the road through which one
>     came into this place".
> 
> I'm leaving to the noexit-on-ways specialist(s) to decide of the exact
> usage/definition and of the contents of its paragraph.
> 
> But please, no more or a minimum of discussion about noexit-on-nodes
> which stands very logical in its shoes.
> The only thing I have to say is that its new rendering shows at one end
> of a road the road sign that's in reality at the other end and that this
> doesn't help much the on-a-way-sister-spec to show that direction, if
> used jointly.
> 
> Is everybody happy now?

Thanks, for your work.
One minor issue is the paragraph about "When not to use", in particular
the advice to use highway=turning_circle instead. It is OK to mention
turning_cirles but noexit might still be useful and there should be no
exclusion for this situation.


Now, we have two different meanings but I still wonder if we need to tag
the second one on ways or if we should state that this meaning is not
really useful and regarded deprecated and only advice to tag noexit=yes
on nodes.

Another solution which came to my mind is to deprecate noexit completely
and use no_exit as replacement and make sure this one is only valid for
nodes.

Cheers fly



More information about the Tagging mailing list