@JassKurn: several of us have found that mapping landuse farmland (and pasture) on a field by field basis is probably the best way of doing it. Things like hedgerows, fences etc are useful particularly to provide detail for walkers so many of the elements will eventually be added at this level too. This has the advantage that the elements are easier to work with & inspect.

tl;dr: no harm done I think

Jerry aka SK53

Hi DaveF,

I've come across this, it looks to be an incorrect preset in the iD editor, clearly people have used landuse=grass to get greens to render, but that is wholly superfluous now. I'm not sure if anyone has raised an issue on GitHub about this, but it needs to be done. It is certainly tedious to have to correct landuse=grass to surface=grass each time & then have the editor yell when you save the edits.

Jerry

Suggest historic=* and or disused:military=*. There are loads of old military sites mapped. The basic tagging for gun batteries is here bunker_type=gun_emplacement (not sure how one deals with ones in use & one's not in use, but suspect vast majority are latter).

Whilst I appreciate that iD often makes somewhat erroneous suggestions, as an experienced editor I would expect you not to accept them as gospel.

I note that "building:type" was only prominently marked as deprecated on the wiki in 2019 not in 2013 when I used the tag. You just cannot assume that: a) someone is aware that a tag has been deprecated by some group of contributors (often a small group); and b) that the tag was not added before anyone suggested it was deprecated. Always the goal should be to leave OSM data no worse than before one edited.

In practice I haven't actually correctly identified the Crane houses: there are very similar houses built almost contemporaneously in brick (but also rendered). There is a plan in this document (I have seen originals btw) https://committee.nottinghamcity.gov.uk/Data/Wollaton%20and%20Lenton%20Abbey%20Area%20Committee%20(Area%207)/20110215/Agenda/SuttonPasseysDraftAppraisalV1%20-%2045933.pdf, and once you know the difference it's easy to check on Mapillary as I have traces on most of the roads.

The building:design suggestion is a good one, and will work better than building:type. I do note the key is not documented & has low usage. To date I have used SK53:NCC_building tags as placeholders (partly because I don't know the original design numbers), see osm.wiki/Nottingham/Buildings/Council_Housing. Others such as BISF, Tarran-Newland, Wimpey No-fines would fit well with building:design.

Please do me the courtesy of asking why I've tagged something in a particular way, instead of 'fixing' it. I have paritally reverted this changeset because it broke more things than allegedly fixed. The Crane Bungalows were entirely validly tagged as building=bungalow, the additional tag indicated more precisely that they are a particular (very rare) type of unusual prefabricated bungalow. Yes another key might be preferred, but if you don't understand a key value please don't alter it without asking why it is tagged that way. BTW this change was much more substantial than a "minor fix, and should have been recorded as such in the changeset comments.

Except that OS Open Local shows it at the top of Cross Street: certainly seems to move about..

As indeed as shown here https://www.geograph.org.uk/photo/3840285

Looking at the post office locations on will_p's map it might be in the convenience store now (so you put it on the correct side of the road)

agggh, accidentally started on St Michaels near St Fagans (not completely used again to PL3's changeset cf iD)

The post office (which may well be closed) which you mapped in Penrhiwceiber seems to be on the wrong side of the road.

Hi,

I've reverted these edits as they lack plausibility & appear to be just playing around with the edit functionality.

Jerry aka SK53

That looks good enough!

@matt_twam_asi: do you have a link to the PRoW modification order? West Sussex website is not particularly forthcoming. I also found this regarding the section across the old station http://www2.westsussex.gov.uk/ds/cttee/row/row230209i7.pdf

NCN 223 implicitly means bicycle=yes or bicycle=permissive, but some routers may not use this information and may need it on the ways. Leave it for a bit, I'm asking other experienced mappers about this. The issue is that fixing it here doesn't fix it for any other bridleway.

The access is a bit complex here, S of the railway it appears to be legally a public footpath, and the bit through the old station isn't even on the West Sussex Rights of Way map (see Map the Paths which helps find rights of way missing from OSM https://www.mapthepaths.org.uk/?lat=51.05093034480947&lon=-0.36507251053827583&zoom=4&mode=0).

Hi mygrove,

Please dont fiddle with access tags to fix a routing problem with an app. Often apps will be using older data, or may just have bugs in the consumption of OSM tags (highway=bridleway is a common one, particularly if the app is written by developers elsewhere).

A quick check on the RideWith GPS website suggests that ways tagged with highway=bridleway with no specific bicycle access set are ignored. Graphhopper also seem to have this myopic view of bridleways. A UK-based cycle router such as cycle.travel does not.

The correct action would be to ask RideWithGPS to fix this bug in their access rules. I've sent such a comment through to them via their website.

When you change the access you can affect 10s of apps many of which may have been working perfectly. The app you are using may not update for another month or so, so you wont even see if you have fixed an issue or not. In particular it is now impossible to do any diagnosis of your original problem, because you have potentially affected routing for everyone (as well as caused the way to be virtually invisible on the main map). I have therefore reverted your changes so that we can really work out what the problem might be (which is probably adding bicycle=designated to the ways).

Sorry to be a bit shirty, but I seem to have spent an hour on this & I'm a bit disgruntled!

Yours,

Jerry aka SK53

Actually I checked further on Robert Whittaker's site and he has an additional page on PRoW data for each authority. He gives Leeds a green tick https://osm.mathmos.net/prow/open-data/. So it might be worth asking him to add Leeds to his progress page &, as you suggest, pinging, Nick Whitlegg too.

Rowmaps has footpaths, bridleways etc as separate geojson files which can be pulled into the iD editor as a vector layer. I'm just going to try to see if it can cope with the footpaths file.

Hi dan000,

I think you need the access tags and opening_hours is much less likely to be consumed. Currently Graphhopper routes through for cycles, cars & pedestrians, but OSRM does not. You probably also need to see how old the Kurviger data is (many routers can take a month or so to update with new data).

Jerry aka SK53

(brought here via Help @OSM)

Hi James,

At the moment the Leeds data does not look to be available in a form which is usable. There is a tool for comparing rights of way mapped in OSM with data we can use here: https://osm.mathmos.net/prow/progress. However, it is available on Leeds Data Mill under OGL https://datamillnorth.org/dataset/leeds-public-rights-of-way.

Robert (rjw62), who maintains it, has great expertise in winkling data (& suitable licences) out of councils, but it takes a great deal of time. It might be worth asking him about the current status (Nottingham also publish PRoW data in this way but I recall there was an issue regarding the license).

One of the interesting things about the Leeds data is that they also provide information on path furniture (stiles, gates etc).

It's also worth noting that until one has walked the line the state of the path might not be clear. I had to file 3-4 issues when I mapped 3 footpaths near Harewood two years ago.

Jerry

Can't believe I just linked to this & you'd checked it's still there! Think I checked 2 years ago, but didn't add a check date.

@eteb: basic multipolygons should be as convenient as drawing the ways selecting each of them and pressing "c". Not sure if you are doing something different.

It gets much more complicated if you cut an existing area as iD converts that to a multipolygon, so generally I avoid it & do things which require it in another editor.