<div dir="ltr"><div dir="ltr">On Fri, 24 May 2019 at 22:12, Kevin Kenny <<a href="mailto:kevin.b.kenny@gmail.com">kevin.b.kenny@gmail.com</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
Yeah, there really are combinations around here:<br>
<br>
does it have signs?<br>
does it have traffic signals?<br>
does it have specific pedestrian-facing traffic signals? (Some<br>
intersections just have you cross at the same time as motor traffic in<br>
your direction rolls)<br>
are the traffic signals pedestrian- or cyclist-controlled? (Is there a<br>
button for you to push?)<br>
does it have pavement markings?<br></blockquote><div><br></div><div>Some of those can probably be simplified away. Like the push button. It may</div><div>seem like a major difference but in actuality on some crossings the ONLY</div><div>purpose of the push button is to light the sign saying "Wait" and the crossing</div><div>cycle is determined by some combination of timing and traffic flow.</div><div><br></div><div>I'd say that traffic/pedestrian signals is the key factor for crossing=traffic_signals,</div><div>irrespective of road decoration even if that road decoration modifies the meaning</div><div>of the signals in some way (it's effectively no different from a sign on a pole).</div><div>A marked crossing doesn't have traffic signals. An unmarked crossing doesn't</div><div>even have markings.<br></div><div><br></div><div>Pavement markings, tactile pavements, dropped kerbs, etc are all attributes. They</div><div>don't turn it into a different type of crossing or (except possibly in Poland) affect the</div><div>interactions between pedestrians and motorists. Nice to map, but as a clarification,</div><div>not a primary feature.<br></div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I'm fine with leaving crossing=* as it is for legacy compatibility,<br></blockquote><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
but we *do* want to move toward orthogonality, since that's what we've<br>
got on the ground.<br>
</blockquote><div><br></div><div>I'm not yet convinced there's orthogonality in crossing type (except possibly in Poland).</div><div>A crossing where the lights mean one thing and the road markings mean a different</div><div>thing doesn't strike me as being even remotely workable: the road markings tell the</div><div>pedestrians they have right of way irrespective of the lights and a green light tells</div><div>the motorist he has right of way. That's no way to run a crossing.<br></div><div><br></div><div>What we may need to do is expand on crossing_ref (maybe with a different name) to cope</div><div>with all the regional differences. "This is a crossing controlled by lights which just happens</div><div>to have zebra stripes, but those stripes do not carry any legal meaning and are purely</div><div>decorative" We almost certainly do need to distinguish between Pelican and Puffin crossings</div><div>in the UK because, although they look almost identical, the light sequences and regulations</div><div>differ. Etc.</div><div><br></div><div>-- <br></div><div>Paul</div><div><br></div></div></div>