I was initially expecting any “was there; no longer there” routes to be in the extinguished table. But I guess the best course of action depends on whether these routes are really extinguished or have just been (accidentally) omitted from the latest PRoW snapshot.
If genuinely extinguished, then they should be removed from the parish tables and added to the extinguished table. If accidentally omitted, then they should be kept in the parish tables. But, of course, the difficulty is knowing which is actually the case as this requires searching out modification orders.
Could you perhaps keep the route in the parish table if currently mapped? Pro: it prevents unknown RoW numbers for mapped routes. Con: no longer highlights potentially extinguished routes and also means some inconsistency in how your tables work.
Or instead, how about a “potentially extinguished” table alongside the “unknown RoW” table, or even just a “potentially extinguished” flag in the “unknown RoW” table?
Apologies for my impatience! As for TR10 104, for some reason it did not occur to me that it had moved, so—having known exactly where it used to be—I simply double checked that location to confirm that it definitely wasn’t there any more and didn’t think to look round the corner. I’ll pop out and map the new location when I can.
Just to let you know, there’s is now a new set of data (from November 2024) in my Post Office comparison tool. You might like to have a look at the comparison maps for your local area to see if anything needs updating.
Also, the “Banking Hubs” that have started appearing in smaller towns around the country are also operated by the Post Office. They’ve included these Banking Hubs in their Post Office branch dataset. They’re not Post Offices, so I’ve stripped them out of the import to the above tool. But I have created a new set for them in my Chains tool: https://osm.mathmos.net/chains/Q131824197/. Any missing or unexpected branches will now appear in Survey Me!.
There’s been a few changes to parishes, with new/renamed parishes appearing for Birdlip, Great Oldbury, Northway, Wormington and Wheatpieces. I think I’ve updated the prow_ref numbers for any existing Rights of Way now in those parishes, but I might have missed some.
I’ve got a couple of questions on the schools front.
Is there still a page on mathmos.net where we can see when each dataset was last updated and processed?
On the “OSM Schools Progress in the ‘X’ Postal Area” pages, there are always some schools from the neighbouring postcode. On the BT postcode page - Northern Ireland - there are therefore lots of schools that are in the Irish Republic, and therefore outside the remit of the Mathmos tracker. Thanks to the recent Irish buildings effort there are now scores of them, none of which can be processed using your tooling. Would you consider filtering them out more firmly?
Thanks as ever for all your work on this brilliant resource!
There’s a dashboard at https://osm.mathmos.net/updates.cgi, which shows when the various data sources were last updated. (And yes, the NI schools source data could do with being refreshed - it’s on my list to donow done.)
With a bit of Overpass magic, I think I’ve managed to strip objects that are in the Republic of Ireland. It seems to work for the Schools OSM data, I’ll try to deploy the same fix to my other tools over the next few days. Thanks for the push to fix this .
HI Robert, thank you so much for the work you do. I’ve been doing a lot of work on missing postcodes in my area using your site when i realised it doesn’t seem to be updating from OSM data, listing 1/1/1970 as the last updated date for that data.
Sorry, that bit of text is out of date. I used to be able to do daily updates of numbers from a taginfo feed, but a while ago taginfo had an update that meant I could no longer do this. All the OSM-derived data is just updated (roughly) weekly now. I’ve just corrected the text on that page to reflect that.
Hopefully, it was just the text that was making you think things weren’t updating. If there’s any actual data that you think is behind, then do let me know.
It looks like they’ve been adding ref=* tags with “Footpath” etc spelt out, alongside prow_ref=* tags with it abbreviated. See e.g. Way History: Great Dunmow Footpath 48 (98596949) | OpenStreetMap . I’m not sure both forms are needed, and ref=* presumably shouldn’t be used unless its signed on the ground.
Robert, I found your tools last night (specifically, the Post Offices one) - it is BRILLIANT
Super useful for data sets like this where, yes, there is a canonical data set, but simply importing onto OSM would create a complete mess, because (i) a not-insignificant amount of the objects are already on OSM just by people adding them organically, and (ii) the position accuracy of the data set is a bit dubious. (Also in the case of Post Offices (iii) when a branch has it’s own building, it makes sense to tag the whole building, but when it’s a PO counter inside another shop, it makes more sense to map it as a node…)
Anyways, I am probably rather late to this party but THANK YOU!
I am going to try and get Northern Ireland’s Post Offices and postboxes looking a bit more respectable.
Here is a suggestion of another tool, which would be extremely useful in Northern Ireland for similar reasons to the above:
NI’s state-owned public transport operator, Translink, does publish its list of bus stops as Open Data Translink Bus Stop List - Dataset - Open Data NI (14650 of them, to be exact).
But, even regardless of any licensing quibbles, bulk importing their bus stop data would make a helluva mess: a modest pecentage of them are already mapped, and the positional accuracy of Translink data is awful (usually out by at least 100 metres - which in practical terms, usually means their data shows a bus stop in the middle of a field or on the wrong street corner etc etc).
A tool to visualise which bus stops from their data are currently missing from OSM would be a godsend, allowing them to be manually added with the benefit of local surveys to determine where they really are…
@rskedgell here is a novel tagging issue - your Post Office tool flagged this one: Node: J Rea Post Office (8825064740) | OpenStreetMap - it’s not on the Post Office Limited list.
Turns out that’s because it’s in the Ulster Folk Museum, a recreation of a 19th century post office Not sure how we tag that
.
