On Thu, Jan 28, 2010 at 12:33 PM, Greg Troxel <span dir="ltr"><<a href="mailto:gdt@ir.bbn.com">gdt@ir.bbn.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div><div></div>If a user deletes a node and replaces it, then the next import should<br></div>
flag that as needing manual intervention. And users shouldn't do that -<br>
they should edit the node.<br></blockquote><div><br>If I find it more convenient to delete a node and recreate it, then that's what I'm going to do. I'll copy the address information to the future node, though. And I'll also keep the "name=Holiday Inn" or whatever. So you might not even need to do any manual intervention.<br>
<br>Download all current versions of all imported nodes. Flag any node with a changed address or changed name for manual intervention (possibly being lenient about name changes such as "Holiday Inn" <-> "Holiday Inn Express"). Download all the nodes with "name=Holiday Inn" (or with a name beginning with "Holiday Inn"). Match them to your database using the address and/or lat/lon. Flag anything off by more than say 50 meters for manual intervention. Make sure every holiday inn found by one of those two methods is included in exactly one node. Flag duplicates and extraneous nodes for manual intervention. For missing nodes, either re-add them if you're certain they are correct or flag them for manual intervention if you're not 100% certain about your data.<br>
<br>In any case, we're only talking about 1,338 locations. The number of items being flagged for manual intervention is likely to be small once the initial import is made.<br></div></div>