Hi,<br><br><div class="gmail_quote">2012/5/16 Alan Mintz <span dir="ltr"><<a href="mailto:Alan_Mintz+OSM@earthlink.net" target="_blank">Alan_Mintz+OSM@earthlink.net</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Tracing coastlines that vary over time and are of unverified positional accuracy, and then stretching those out to cover the supposed territorial limits (and knowing the correct methods and exceptions) seems like the wrong answer here. It really seems these should come from an import of the officially defined and negotiated international boundaries, no? There were some issues with claimed (perhaps invalidly) copyrights over some of the UN data, but one of those UN datasets should be available by now. Or perhaps a US government has a copyright-free dataset for us?<br>
</blockquote><div><br>Fully agreed. Tracing coastlines is definetly not the ideal situation - although perhaps the best we can get for the time being. The resulting error is probably in the same range as we see with imported borders ie. from WDB as those are sometimes pretty generalized.<br>
<br>In the corresponding act I linked to in my previous post, Tonga refers to British Admirality maps as the reference for the coastline to define the baseline. So I doubt that there's a legal way from those maps into OSM...<br>
<br>There is a database of maritime boundaries [1] but I think the license is not compatible with OSM. But perhaps it might be worth the effort to ask them directly.<br><br>Michael <br><br>[1] <a href="http://www.vliz.be/vmdcdata/marbound/">http://www.vliz.be/vmdcdata/marbound/</a><br>
</div></div>