For more information see changeset/139189031 , way/553974997 and https://en.wikipedia.org/wiki/en:Threecastles%20Castle .

> I can see how annoying you must find it to interfere with what you would like your map to look like.

No, that is not the issue at all. The issue is that you are removing information that other people have added to OpenStreetMap. The keys and values processed by that map style are based entirely on the keys and values present in OSM; they are a reflection of the richness of data present in OSM. If someone removes data from OSM (as you did here) any map consuming OSM data (including mine, but also including many others) will be poorer for it.

> So how about adding fortification_type=castle?

That would get us back to where we were (and was a combination in use for this sort of feature previously). However, given what we now know about the source and relation between these objects here, maybe a site relation would also make sense?

Also - (1) are you going to make this change or do I need to, and (2) what about any other similar tagged objects that you removed information from in the same time?

> Also, if you check the official OSM wiki on notes,

It says "Don't use notes for yourself in a way which is useless to others. Although you can use notes as a reminder to yourself, you are also inviting others to look at it. "

You are no more an "official" representative of OSM than any other of the 2 million or so other OSM contributors.

If you want to build a "what if" model based on OSM as well as some other data it's pretty straightforward to do so. I'd suggest asking how to create what you need over at the forum https://community.openstreetmap.org/ .

> This was a value not in use

This is untrue - as you would have seen had you looked at https://taginfo.openstreetmap.org/tags/archaeological_site=castle#projects !

> by the very mapper who set out to restructure the historic=archaeological_site family half a year ago ... Since nothing is left of this castle to be seen (aerial imagery shows a plain meadow), mapping may be questionable in the first place,

Did you ask her about it? I bet she'd have been able to help :)

The tiniest bit of investigation by me found https://en.wikipedia.org/wiki/en:Threecastles%20Castle , which is linked from way/553974997 , which mentions old OS maps, which are available in iD, and this is clearly the "other" castle mentioned in the wiki page and clearly visible on the old OS map.

> There isn’t yet a clearly established tagging scheme for former,

Sure there is: https://taginfo.openstreetmap.org/keys/fortification_type#values

> So I specified this, not removing information, but adding information by being more precise.

That is simply untrue. It was clearly mapped before as "something that used to be a castle". Now it is mapped as "something that is some sort of fortification, but with no information about the period or structure". See https://map.atownsend.org.uk/maps/map/map.html#18/53.18182/-6.48529 for the difference that it causes - now the eastern one is just "some sort of fortification" and the western one "some sort of castle".

Sigh. You're removing information from OSM again. If you look at the history of node/3153973098/history you can see that something allegedly was a castle. You've changed that to just "fortification", of which a (former) castle is just one kind. There's at least one documented data consumer (me) that distinguishes between different historic and archaeological fortifications.
You may, of course, have rwsearched exactly what this site is/was - if so, I'd be delighted to know what you've found....

Also, in what way is it "official"?

Please don't add personal note data like this. See osm.wiki/Verifiability .

Openstreetmap is a map of the world as it is, not as we'd like it to be.

Thanks!

For info, this seems to have accidentally added a large fence to an area of heathland. See https://community.openstreetmap.org/t/1-member-multipolygons/101507/84 when I happened to spot it in response to a question about something else.

Agreed - some of these changes really do need to be discussed

Hello,
I wonder if you know which way the NCN66 cycle route goes across the gap at relation/17883#map=18/53.78587/-0.36737 ?
Cheers,
Andy

Thanks for mapping these, but assuming that they are actually square, you can square up these irregular quadrilaterals by pressing "q" in most editors.

For info I've commented on changeset/131478650 about the merging and the currently silly elevant value.

You seem to have merged the peak https://osm.mapki.com/history/node/321811440 and the trig point https://osm.mapki.com/history/node/5317253878 here in error . It's a while since I've been there, but I don't remember the trig point being right at the very top, but offset to one side. The previous elevation values were 931 and 931.2, which would make sense. The current elevation of "931;931.2" obviously does not.

Thanks - is the accountant upstairs or something?

Actually - not this one (I was thinking of one to the north), but I'm sure that it's not "shop=vacant".

Hello, is https://osm.mapki.com/history/way/961287358 still a clothes shop? I'm guessing not...

Hello, I'm guessing that shop=art was left over here? See https://osm.mapki.com/history/way/204019352 . I've removed it, hope this is OK!

OK - thanks!
I'll update my "check list" https://github.com/SomeoneElseOSM/database_qa_scripts/blob/main/osm_ldp3#L119 based on the current contents of the superroute relation/1992599 .
I'm pretty sure that there is more work to do around Hull - relation/2452530 is the new route of NCN1 (and is incomplete) and the bit that goes to Hull and back at relation/15975674#map=12/53.7664/-0.4168 is incorrect, though those are part of other routes. chillly, who is local to Hull and the author of relation/2452530 , may be able to advise further.

I agree with Kai here - it doesn't make sense to remove information like this without discussion with other mappers. After that discussion it's possible that better suggestions will emerge, but "tagfiddling" like this doesn't improve the quality of OSM data one iota.
- Andy (writing, for the avoidance of doubt, in a personal capacity)