Looking good...I don't see any missing ways after an initial run and a quick glance. I see that you've listed all of the TLID's when merging stuff, so I can see now how it provides a path to import address data later.
<br><br>-Al<br><br><div><span class="gmail_quote">On 7/13/07, <b class="gmail_sendername">Dave Hansen</b> <<a href="mailto:dave@sr71.net">dave@sr71.net</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Changes from 0.3:<br> - Fixed lots of ways getting dropped (I was too aggressively reclaiming<br> objects that I thought I was done using)<br> - I used to keep hashes of arrays of ways with each name and tried to<br> find ways to glob together by searching. Trying to combine two ways
<br> together turned out to be one of the CPU dominating parts. Now, it<br> indexes all ways by starting and ending node. Much more efficient<br> to find ways to join together.<br> - started including some tiger identifiers in the Node tags. This
<br> should let us use the tiger data at a later date to add addressing<br> information.<br><br>Using the name-indexed method.<br> combine_fail: 956461<br>combine_success: 2612<br><br>Using the start/end node in a way method:
<br> combine_fail: 6181<br>combine_success: 2669<br><br>If JOSM complains about running out of memory, increase the Java heap<br>size by running it like this:<br><br> java -Xmx1024M -jar josm-latest.jar tiger-file.osm.xml
<br><br>See the README for detailed instructions.<br><br><a href="http://sr71.net/~dave/osm/tiger/tiger-to-osm-0.4.tar.gz">http://sr71.net/~dave/osm/tiger/tiger-to-osm-0.4.tar.gz</a><br><br>-- Dave<br><br><br>_______________________________________________
<br>dev mailing list<br><a href="mailto:dev@openstreetmap.org">dev@openstreetmap.org</a><br><a href="http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev">http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev
</a><br></blockquote></div><br>