<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Does this deserve a lightning talk or breakout session next weekend?</div>
<div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
--</div>
<div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Andrew<br>
</div>
<div id="appendonsend"></div>
<hr style="display:inline-block;width:98%" tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" style="font-size:11pt" color="#000000"><b>From:</b> Cj Malone <me-osm-talk-gb@keepawayfromfire.co.uk><br>
<b>Sent:</b> 01 July 2021 23:39<br>
<b>To:</b> talk-gb@openstreetmap.org <talk-gb@openstreetmap.org><br>
<b>Subject:</b> Re: [Talk-GB] UPRN to postcode lookup</font>
<div> </div>
</div>
<div class="BodyFragment"><font size="2"><span style="font-size:11pt;">
<div class="PlainText">I believe this data to be sufficiently open for use in OSM. While I<br>
don't believe they got the postcodes from Code-Point Open, they do<br>
mention that they worked with OS to insure no postcodes were released<br>
that isn't included in Code-Point Open. I believe both OS and ONS are<br>
fully consciously publishing uprn -> postcode under the OGLv3 and that<br>
this isn't one of the OGL "open but not" gotchas.<br>
<br>
While I hadn't considered Jerry system to infer house numbers, this is<br>
an important point as now the vast majority of an address is Open.<br>
<br>
OS Open UPRN gives us locations.<br>
OS Open USRN, OS Open Names and OS Open Linked Identifiers give us<br>
street names. (Usually, there is some quality issues here)<br>
ONS gives us a perfect postcode. (nolonger, educated guesses from Code-<br>
Point Open)<br>
City/Local authority can also be calculated.<br>
<br>
As Jerry also said, we can trim OS Open UPRN by buildings from OS Open<br>
Local to get a reasonable list of properties, skipping most sub<br>
stations, roads, post boxes, phone boxes and other UPRNs. This isn't<br>
perfect some property UPRNs are outside the OS Open Local building<br>
footprint, and some others will be inside.<br>
<br>
I'd also trim OS Open UPRN to exclude points from within OSMs<br>
landuse=commercial. UPRNs in these areas are to dense to be understood,<br>
especially at this time.<br>
<br>
I'd also trim UPRNs that share locations (apartments), and possibly<br>
those that are "extremely" close.<br>
<br>
Tying in with with your address project, we could present these points<br>
to users, allow them to add the house number and/or name. As well as<br>
verify the street. Then add that point to OSM.<br>
<br>
I'd argue to also get that data available to StreetComplete so it can<br>
be added on surveys, that would mean the data would then again have to<br>
be filtered against OSMs buildings and existing addresses.<br>
<br>
See this StreetComplete issue for use third party data.<br>
<a href="https://github.com/streetcomplete/StreetComplete/discussions/2477">https://github.com/streetcomplete/StreetComplete/discussions/2477</a><br>
<br>
I also wouldn't be against adding these address points to OSM before we<br>
get the house number, or we trace the building. It would certainly<br>
enable us to rapidly expand out postcode coverage.<br>
<br>
Cj<br>
<br>
<br>
<br>
_______________________________________________<br>
Talk-GB mailing list<br>
Talk-GB@openstreetmap.org<br>
<a href="https://lists.openstreetmap.org/listinfo/talk-gb">https://lists.openstreetmap.org/listinfo/talk-gb</a><br>
</div>
</span></font></div>
</body>
</html>