<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">2014-08-14 12:24 GMT+02:00 André Pirard <span dir="ltr"><<a href="mailto:A.Pirard@ulg.ac.be" target="_blank">A.Pirard@ulg.ac.be</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br><div bgcolor="#FFFFFF" text="#000000">
<div>On 2014-08-14 11:08, Janko Mihelić
wrote :<br>
</div><div class="">
<blockquote type="cite">
<div dir="ltr">
<div>Well first, tunnel=yes is obviously wrong. We need to
replace this with cave=yes. Other than that, I have no
problems with this. If a cave has two cave entrances, then
information that they are connected by footpaths is valuable
information. <br>
<br>
</div>
Janko<br>
</div>
</blockquote></div>
Obviously? Regarding paths and waterways, especially ones fitted up
for tourism, I wonder...</div></blockquote><div><br></div><div>Ok, I exaggerated with the word obviously. But tunnels are used a lot in OSM and they have gotten a very clear semantic meaning: a man made opening in the ground. There's no reason to add caves to that definition.<br>
<br></div><div>The reason "renderers don't render it" is invalid because of one of the oldest rules in OSM: don't tag for the renderer.<br><br></div><div>And if a data consumer tries to find the length of all footpaths in tunnels in Poland, she will ask "find all highway=footpath + tunnel=yes", and she will find all tunnels plus all caves. And that isn't right.<br>
<br></div><div>I suggest starting the cave=yes tag, and we'll see, maybe the renderers will pick it up. They only have to treat it the same as tunnel=yes.<br></div><div><br></div><div>Janko</div></div></div></div>