<div dir="ltr"><div dir="ltr">On Sat, 10 Aug 2019 at 17:27, Julien djakk <<a href="mailto:djakk.geographie@gmail.com">djakk.geographie@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>
The "old" highway tag can give default values to the 5 new tags, so it<br>
is not necessary to re-map everything :)<br></blockquote><div><br></div><div>If it is a guaranteed 1:1 relationship then there is no point doing it. Maybe if the</div><div>semantics became clearer (like migrating landuse=grass to landcover=grass)</div><div>but that doesn't appear to be the case.</div><div><br></div><div>But your scheme appears to offer much more details and more precision. So</div><div>you might say that highway=primary automatically becomes a=x + b=y + c=z</div><div>but that implies that a highway=primary might actually be a=x + b=y + c=w.</div><div>So it is necessary to remap.everything in order to check whether the primary</div><div>highway has c=z or c=w. It might not be urgent if there's no great practical</div><div>difference between c=z and c=w, in which case there isn't much need for</div><div>the new scheme.<br></div><div></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<br>
Yes you are absolutely right : I need my own renderer to populate the<br>
new tags. I was thinking about putting anything (roads, summits,<br>
footways, towns, trees …) with importance = 1 to the lower zoom, etc.<br>
(Actually tag importance already exists, used for railways, has main<br>
values regional or national).<br></blockquote><div><br></div><div>Then you not only need your own renderer, you need your own database. <br></div><div>Two of them, in fact. One to hold a regularly-updated copy of the OSM</div><div>database and one holding your extra tags. Because you'll probably</div><div>get a lot of opposition to the idea of you shoving your own tags into</div><div>the OSM database for no reason than that you like the idea even though</div><div>nobody else does.</div><div><br></div><div>People have managed to get their own tagging sub-schemes accepted: the</div><div>sea mark and historical objects tags come to mind. A lot of their stuff isn't</div><div>rendered by anything other than their own carto (some things are, but most</div><div>are not). BUT those tags aren't applied to every object in the database.</div><div><br></div><div>Oh, and there's the problem that "importance" seems, at this stage, to be a</div><div> subjective value decided by you but applied to everything. Even though for</div><div>many practical purposes renderers make their own decisions about what</div><div>to display at different zooms, you wish to impose your view on everything.</div><div><br></div><div>The more you try to justify this idea, the less feasible it seems. The first law</div><div>of holes comes to mind...</div><div><br></div><div>-- <br></div><div>Paul</div><div><br></div><br></div></div>