<div dir="ltr"><div><div>My fault.<br></div>Loops (in the sense of self-intersections) are not a problem. Drop rule B(2)!<br></div>What I had in mind was roundabouts. They are not loops. Thy are special cases of the rule B(3)<br></div><div class="gmail_extra"><br><div class="gmail_quote">On 3 May 2018 at 19:10, marc marc <span dir="ltr"><<a href="mailto:marc_marc_irc@hotmail.com" target="_blank">marc_marc_irc@hotmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">"impossible" only due to missing tools to be able to do the job<br>
and the time it take do it "by hand"<br>
try to order this one <a href="http://www.openstreetmap.org/relation/1744381" rel="noreferrer" target="_blank">http://www.openstreetmap.org/<wbr>relation/1744381</a><br>
I spent a little time sorting these members, but it's a waste,<br>
it's deteriorating faster than it's getting better.<br>
<br>
Le 03. 05. 18 à 18:49, Colin Smale a écrit :<br>
<div><div class="h5">> Why does a loop make it impossible to sort the ways? It implies that a <br>
> section of the route is present twice in the relation, but there is <br>
> surely no distinction between the first traversal of a way and the <br>
> second traversal?<br>
> <br>
> <br>
> On 2018-05-03 18:42, Volker Schmidt wrote:<br>
> <br>
>> I will try to explain this in a more systematic way:<br>
>><br>
>> Routes belong to either of two categories:<br>
>> (A) Those whose members can be sorted into a single ordered sequence<br>
>> (B) Those that cannot be sorted into a single ordered sequence of members<br>
>> Sorting makes only sense for category (A)<br>
>> Routes of type (B) can be subdivided into routes of type (A), each of <br>
>> which can be sorted, but the overall route can not be sorted.<br>
>><br>
>> Routes are of type (A) if<br>
>> (1) the path from begin to end is identical to the reverse path with <br>
>> all members traversed in the reverse order and in the opposite direction<br>
>> or<br>
>> (2) all members have the role forward<br>
>> or<br>
>> (3) all members have the role backward<br>
>><br>
>> Any route<br>
>> (1) that has more than two ends<br>
>> or<br>
>> (2) that contains any loop (except the case that the entire route is a <br>
>> single loop)<br>
>> or<br>
>> (3) that contains any element with role forword or role backward <br>
>> (except the cases of all-forward or all-backward)<br>
>> or<br>
>> (4) that contains node or area elements<br>
>> is of type B<br>
>><br>
>> I am not sure if I have taken care of all cases - please complete as <br>
>> necessary<br>
>><br>
>><br>
>><br>
>> ______________________________<wbr>_________________<br>
>> Tagging mailing list<br>
</div></div>>> <a href="mailto:Tagging@openstreetmap.org">Tagging@openstreetmap.org</a> <mailto:<a href="mailto:Tagging@openstreetmap.org">Tagging@openstreetmap.<wbr>org</a>><br>
>> <a href="https://lists.openstreetmap.org/listinfo/tagging" rel="noreferrer" target="_blank">https://lists.openstreetmap.<wbr>org/listinfo/tagging</a><br>
<div class="HOEnZb"><div class="h5">> <br>
> <br>
> ______________________________<wbr>_________________<br>
> Tagging mailing list<br>
> <a href="mailto:Tagging@openstreetmap.org">Tagging@openstreetmap.org</a><br>
> <a href="https://lists.openstreetmap.org/listinfo/tagging" rel="noreferrer" target="_blank">https://lists.openstreetmap.<wbr>org/listinfo/tagging</a><br>
> <br>
<br>
______________________________<wbr>_________________<br>
Tagging mailing list<br>
<a href="mailto:Tagging@openstreetmap.org">Tagging@openstreetmap.org</a><br>
<a href="https://lists.openstreetmap.org/listinfo/tagging" rel="noreferrer" target="_blank">https://lists.openstreetmap.<wbr>org/listinfo/tagging</a><br>
</div></div></blockquote></div><br></div>