<br><br><div class="gmail_quote">On 3 March 2011 10:14, Jerry Clough : SK53 on OSM <span dir="ltr"><<a href="mailto:SK53_osm@yahoo.co.uk">SK53_osm@yahoo.co.uk</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div bgcolor="#ffffff" text="#000000"><div class="im">
On 03/03/2011 09:51, Richard Mann wrote:<br>
<blockquote type="cite">
<pre>I think you might also consider a path density map or a
shop/pub:street density map. That's the sort of stuff where OSM can
really do much better than OS / Google.
Richard
</pre>
</blockquote></div>
This was exactly what I was trying to do with my various <a href="http://sk53-osm.blogspot.com/2011/02/updating-pub-density.html" target="_blank">pub
density maps</a>. I did do various attempts to normalise pub
density to highway length but none of them showed anything
interesting. I have also looked at turn restrictions, and public
toilets as potential proxies for things mapped on the ground versus
things mapped remotely. A recent diary entry pointing to Gregory
Williams cycle-parking heatmap highlighted another possible
candidate. Unfortunately most of these maps (like the Botanical
Society's <a href="http://www.bsbimaps.org.uk/atlas/map_page.php?spid=2952.0&sppname=Cotoneaster%20rehderi&commname=Bullate%20Cotoneaster" target="_blank">maps
</a>of distribution of <i>Cotoneaster </i>species) mainly show
where mappers live or are active.<br>
<br>
As more of the highway network gets completed by remote mapping, the
more important it is to find handles for on the ground mapping.<br>
<br>
Pete Reed did some <a href="http://tlatet.blogspot.com/2010/04/more-osm-coverage.html" target="_blank">nice
comparisons</a> between highway length by authority as reported by
the DoT and OSM road length. I'm not aware of him having updated
these recently.<br>
<br>
Incidentally, 0% discrepancy between OSM and OS Locator is
inadequate as an indication of streetname completion: the next test
would be to check OSM names against PAF to see how many address
elements were missing. Unfortunately, we would not be able to use
any of the detail: but even headline figures by LA might be
interesting. I've noticed a trend in new in-fill huosing
developments for houses facing a main road to have a separate name
(e.g., <a href="http://osm.org/go/eu8Z4HWl1--" target="_blank">here</a>), which
does not appear in Locator, nor do side terraces in
late-19C/early-20C housing (e.g., <a href="http://osm.org/go/eu8ZmJWOR--" target="_blank">here</a> or <a href="http://osm.org/go/eu8Zn0fu1--" target="_blank">here</a>). All these examples
are places I've added this year.<br>
<br></div></blockquote><div><br>I agree that the whole concept of 'completeness' is problematic but I is a very useful way-point in that direction. As far as current OS data users that is what they will think of as 'good enough' so for them it is a useful concept.<br>
<br>I do want to do analysis of how many additional names we have that are not in OS Locator. Possibly we call that 105% OS Locator completeness!<br><br><br>Regards,<br><br><br><br> Peter<br><br></div><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div bgcolor="#ffffff" text="#000000">
Jerry<br>
<br>
</div>
<br>_______________________________________________<br>
Talk-GB mailing list<br>
<a href="mailto:Talk-GB@openstreetmap.org">Talk-GB@openstreetmap.org</a><br>
<a href="http://lists.openstreetmap.org/listinfo/talk-gb" target="_blank">http://lists.openstreetmap.org/listinfo/talk-gb</a><br>
<br></blockquote></div><br>