Hi, Graham. Just for wider knowledge, I'm doing some work at the moment, along similar lines, but in support of an app, not a bulk import.
Stating the obvious, before any bulk imports are considered, the source data licensing would need to be checked. I suspect people have been round this bouy a few times, so I'll step back, now. Chris chris_debian On Tue, 9 Jun 2026, 12:01 , <[email protected]> wrote: > Send Talk-GB mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.openstreetmap.org/listinfo/talk-gb > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Talk-GB digest..." > Today's Topics: > > 1. Revisiting OS trigpoints - viability of importing? () > > > > ---------- Forwarded message ---------- > From: Graham Whaley <[email protected]> > To: [email protected] > Cc: > Bcc: > Date: Tue, 9 Jun 2026 11:55:01 +0100 > Subject: [Talk-GB] Revisiting OS trigpoints - viability of importing? > Hello, > first time poster here. I've done a very few small submissions to OSM in > the past (user gwhaley), and just restarted updating small items around my > location (having spotted some new bins and benches etc. appearing). > > During that work I realised the trigpoints near me are not in OSM, which > led me down the path to this very useful wiki page from @gregrs around 2016: > https://wiki.openstreetmap.org/wiki/Ordnance_Survey_triangulation_stations > and a small related discussion on the list around 2019: > https://lists.openstreetmap.org/pipermail/talk-gb/2019-July/023176.html > > This inspired me to grab the current datasets and do some crunching. I've > posted my code and findings on github at: > https://github.com/grahamwhaley/OSM_UK_trigpoints > > I'd like to discuss if the community think it is viable to import a bunch > of OS trigpoint data? I think it is. I don't think we could automate every > trigpoint found in the OS data (see below and/or the github page for > details), but I think there is scope to import the vast majority of OS > trigpoints as nodes. Of course any import would follow the path of > following the import guidelines, writing an import plan and > approaching/announcing to the wider community before any automation > occurred. I'd also fully expect to do some dry runs and some smaller region > imports before doing the bulk of the data. I'd also gladly welcome any > help, reviews etc. > > I'll drop a few statistics here... > > - The OS trigpoint database has 6081 'live' pillars > - The OSM database has 2804 items purporting to be trigpoints - 2560 > after some filtering > - The OS benchmark data (which gives us FlushBracket numbers) has 3703 > useable entries > > After some data crunching and filtering (see the github page - but I fully > expect to adapt and refine the current code), we end up with the 6081 OS > nodes being classified into the following groups: > > - 60 OS entries that seem to align with existing OSM entries so likely > need no or little change > - 4440 potential 'new' OSM trigpoint entries - entries from the OS data > that do not seem to align with any existing OSM nodes > - 246 'editable' entries, where an OS entry nearly aligns with an OSM > entry, but we could 'coerce' the OSM entry to match the OS entry > - 1335 entries that could do with human review - the OS and OSM entries > are both similar and different (for example, might have matching names but > different FB numbers), and we could try and figure out why and what to do > in each instance. > > I have some ideas about how to proceed, but obviously would like community > input. My overview thoughts are: > > - Adding the 4440 'new' OS trigpoints feels like the least risk, as it > would not be trying to duplicate or modify any existing OSM nodes > - The 'editable' nodes might be moderately easy to apply after a fairly > thorough human review of the changes. > - The 1335 entries that could do with human review - maybe we can get > them on a > site like/similar to Survey Me or FHODOT to make them accessible for > ongoing community review. > > Many thanks, > Graham > > _______________________________________________ > Talk-GB mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/talk-gb >
_______________________________________________ Talk-GB mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-gb
