<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto"><br><br><div dir="ltr">sent from a phone</div><div dir="ltr"><br><blockquote type="cite">On 30 May 2022, at 22:16, Volker Schmidt <voschix@gmail.com> wrote:<br><br></blockquote></div><blockquote type="cite"><div dir="ltr"><div>Let's take a real-life example.</div><div><a href="https://www.openstreetmap.org/way/79099759" rel="noreferrer noreferrer" target="_blank">This street</a> has "sidewalks" on both sides, but they are <a href="https://www.mapillary.com/app/?pKey=469620874291519" rel="noreferrer noreferrer" target="_blank">not continuous</a>. There are no ramps. When I put sidewalk=both I was really hesitating. </div></div></blockquote><br><div><br></div><div>this one is a typical situation with many crossing driveways and no ramps, still sidewalk=both is correct.</div><div><br></div><div>I have seen other interruptions e.g. with tree trunks growing on the sidewalk or building corners (historic situations) and similar, which reduce the sidewalk width to sometimes less than the width of a wheelchair (you can walk there if you are able, but not while pushing a stroller or bike or sitting in a wheelchair).</div><div>In other cases, the sidewalk surface is extremely deformed (tree roots).</div><div><br></div><div>I am not going to remove the sidewalk presence tags for these reasons, but could imagine adding micromapping tags (depending on the situation something from “obstacle”, “smoothness”, “surface”, “width”). </div><div>Maybe similarly your case could be micromapped with barrier=kerb tags?</div><div><br></div><div>Cheers Martin </div></body></html>