Oops - typing fail above - "some of this obviously planted for forestry, so not obviously" should have been "... some not obviously".

Hi avatar6, with regard to relation/375035/history , how did you know whether to use natural=wood rather than landuse=forest here? On osm.wiki/Forest there are four (!) conflicting approaches documented for tagging wood vs forest - it would be interesting to know which one you were using.

Unfortunately I'm not familiar with this particular wood (though I've done a fair bit of mapping just to the west - in the Llanthony valley and around Abergavenny) so I can't really really say much about this bit beyond what you can see from the aerial imagery (which isn't much - some of this obviously planted for forestry, so not obviously).

Did you think about asking the previous mapper? Although they mapped it a long time ago, they do still seem to be active. When they did map it, the contents of the "landuse=wood" wiki page was as follows: osm.wiki/w/index.php?title=Tag:landuse%3Dwood_%28Don%27t_use%29&oldid=369680 so they can't really be accused of "tagging it wrong" (by the standards in place at the time) either.

OpenStreetMap is a collaborative project; we have to work together to create the best map. Perhaps if you'd talked to the previous mapper you might have been able to find out which bits of it where deciduous and which bits evergreen, and how much of it is still present (what was mapped 6 years ago doesn't quite match the Bing imagery, and likely of course that might be out of date now too).

Finally, as woodpeck says, "fx lu=wood" really doesn't explain what change you made, why, and what the source of that change was :)

Is the name of node/2661890946 really Poultry Farm, or is that just a description of what it is?

I'm guesssing that you got the name from OS OpenData StreetView, and from experience they often use descriptions rather than names on their map ("Poultry Houses" is another one).

I suspect that if this place is still a working farm (which is not at all definite) it'll have a real name rather than "Poultry Farm".

Using "highway=track" with "area=yes" is going to confuse data consumers (it did confuse me).

Here way/378210101 is just a duplicate of way/378216579 , which is a linear feature and not really an area at all.

In some cases it might make sense to tag a the wider area around highway as an area, and in such cases I'd use area:highway like way/361984588 so that there's no confusion between routable highways and the "areas" around them.

In addition to "wider area around a mostly linear feature" there are also pedestrian plazas for which area=yes really does make sense (routers can make do by routing around the edges).

Here however, unless it's vastly changed since I was last there (and the imagery suggests that it hasn't) an area of any sort here is not appropriate. It's just a linear track.

node/2831638186 is a Garmin waypoint that's been converted to a node. You'd need to ask the original author of changeset/22086457 what each symbol meant when they recorded it. Of the tags only ele is likely to be of any use, and even then only if the Garmin was calibrated before recording (so obviously not in this case since ele=-73 is unlikely!).

Thanks for this (and for resolving those notes).

Hi, I checked with someone in the #osm IRC channel who did Welsh as part of their degree, and apparently "y" is the correct form here. I suspect that it's a bit like English placenames such as "Bridge of Allan" - you wouldn't capitalise the "o" of "of".

c2r at note/464953 suggests using carriageway_ref for those, which makes sense to me. They're not names, really, are they? See also osm.wiki/Names#Name_is_the_name_only .

Thanks - I'll try and figure out where to create an issue for "please don't create global changesets by default".

Just to clarify, the example I gave earlier node/502390265 was a _verified_ customary stop (I caught a bus from there). Different areas' NaPTAN data have been of very different quality - some areas have found stop positions to be reasonably accurate (at the time of import, which was a while ago) some definitely not so. In the case of an area well populated with mappers, like the West Midlands, contacting local mappers always makes sense.

Thanks - it was probably intended to be construction, but may actually be complete now, so I've added fixmes in changeset/35186838 . There were actually other tag problems with the other two areas there too.

Thanks for that. As with Cefn y Cnwc, the "Y" appears to have got capitalised here too.

What's the source of the name "London Orbital Motorway" here? See also note/464953 . In particular see http://ris.dev.openstreetmap.org/oslmusicalchairs/map?osl_id=846862 which shows no name on the M25.

What's the source of the name "St. Albans By-Pass" on way/145478061/history ? I've never seen it on a roadsign. See also note/464953 .

You've added names such as "J 21 J" here. What's the source of that data? See osm.org/query?lat=51.7173&lon=-0.3875#map=16/51.7174/-0.3832&layers=N for more details.

Added to OHM in http://www.openhistoricalmap.org/changeset/1901

Just wondering - if it's a hill is it perhaps a "natural=peak" rather than a "place=locality"?

Relevant issue for OSM carto is https://github.com/gravitystorm/openstreetmap-carto/issues/1280 . I'd suggest that an on-the-ground mapper would need to choose what to use instead though.

My recollection was that we went with "physically_present=no" as per node/502390265 for verified customary stops. Certainly many came in from NaPTAN without highway=bus_stop on them (though my recollection was that the west mids NaPTAN import was slightly earlier and may have used different tags). Locally I've added highway=bus_stop where I've verified that a bus stops; other communities may have done something different.

My understanding of Welsh spelling is that "Cefn y Cnwc" is preferred over "Cefn Y Cnwc" (in that you see the first form far more than the second). What evidence do you have that "Cefn Y Cnwc" is correct here?