I know, you're right. 95% of the time I do add a comment, but I've been editing dozens of churches where (usually) the only change is adding the denomination and adding a comment takes longer than the change itself! The Monkleigh change was unusual in that I did the residential area too, so probably should have, but I was on a roll...

Hi again. The responses I had from Talk-GB agree with me. The coastline should follow the MHW which seems most accurate on OS OpenData StreetView. I'll improve it at some point, but either way it seems clear that the 'island' should be marked as something other than coastline.

Hi. My changes weren't really anything to do with the coastline, being just a decision on whether county boundaries should follow high- or low- water levels.

But to answer your question I would argue the following:
i) The main coastline in this area is inaccurate, and should really follow the Mean High Water level.
ii) The MHW on OS OpenData StreetView runs around the 'island' to the north...
iii) ...but the 'coastline' on OS LandRanger maps (which should also be the MHW) runs well to the south of the 'island'.

So I'm confused. I think it's clear that the main coastline needs refining, but it's not clear where to. I might post on OSM-UK to gather opinions.

>95% of the boundaries follow present-day admin boundaries so I use those existing ways. I've been following the ABC's definition which are those on wikishire. I could have taken a dump of wikishire and imported it but that would have duplicated those 95%+ (and the rest largely follow rivers), so I've been adding the relations to existing ways so it's simple to determine where the traditional boundaries DON'T follow existing admin boundaries, as that's where the interest lies.

I agree I need to settle on a single coastline paradigm. I'll have a think.

Hi Colin. I've oscillated between the two, and am not really happy with either the coastline or low water line. In the end I went with the one to mirror the traditional boundaries on http://communities.maps.arcgis.com/apps/Compare/storytelling_compare/index.html?appid=7b0e661ef66b4a7aacb5a9acf55108ac , and https://wikishire.co.uk/map/ (though not till I'd done the whole of the south coast!). It was the fact that the Wirral and Wash just looked all 'wrong' in outline, as every county map I've seen stops at the coastline. But I'm happy to take advice here. Doesn't part of it come down to whether the traditional counties are administrative or not, which everyone seems to agree they're not?

Years ago when I wasn't so familiar with OSM, I tended to copy however things were done in a particular area, and someone had done dozens with both a node and a way. Now, having been told off by others, I go round doing exactly what you say and transferring the tags to the building!

Hi Will,
I didn't have chance to finish the whole of the traditional border today; I'll try to tomorrow. (I started with duplicating the ceremonial border and am editing the parts that have been altered with modern boundary legislation. The edits I'm doing are as to recreate the traditional borders as accurately as possible (ie probably within 10m), using First Edition OS maps where needed. As you can see from https://bit.ly/1hv6ZL5 (somewhat less accurate than the OS maps) 'ceremonial' Nottinghamshire deviates far less from the traditional boundaries than most other counties. The primary reason that the traditional counties persist today is that they aren't changed by the modern boundary tinkering but reflect how they were for the centuries before. I'll correct Ilkeston (and the rest of the Erewash wiggles) as soon as I get chance.

(Re: adding tags to both ways and relations) Are you talking about the boundary=historic on the way? -- every admin boundary has boundary=administrative in both the way and relation. I'm fine to leave those off the way, but isn't the standard to put an indicative tag on the way so potlatch, etc make it clear what it is?

You make a good point, and initially I didn't have one until reading the boundary=historic wiki page. I added the historic counties as they're not really 'historic' but current in a non-administrative sense. Similarly they had administrative function but only historically, so I'm not sure which to go with. Others have already written to me claiming the historic counties have no place on the map (which I strongly disagree with), so I suspect there'll be a full discussion of all this at some point!

In this case they are not 'historic' but current. (Now with governmental backing also: https://www.gov.uk/government/news/by-george-englands-traditional-counties-can-return-to-englands-roads ) They represent the immutable divisions of England from Saxon times until the boundary acts of the 1880s, and so the 'date' of those I'm adding is "all dates prior to the administrative adjustments that began around then". I've thus been using early-edition OS maps.

I agree historic mapping is indeed out of scope, but England's "historic counties" don't fall into that. To quote a (21st-century) map that I own, "our geographic counties have never been affected, nor their boundaries changed by the endless shifting of local government boundaries over the past 125 years, but have remained constant and are today as they have been for centuries."

Hi,
I've been adding them from new as none exist at present. The result of my work thus far is
http://overpass-turbo.eu/s/ARB
Given it's a manual job, I have to frequently go back and "correct" sections I missed or the like, but although I frequently add the relation to existing boundary ways, I of course never edit the nodes of the way itself.

Thanks for that. Is there an easy way of selecting and changing all of them at once? The "select all" only selects those that are loaded in your current view.

The Ordnance Survey only lists the name Chare Bank in Ebchester itself - well before it reaches the bridge. It then lists B6309 after that. In addition it gives the name as Fine Lane when it turns north temporarily. On the straight part from there to Whittonstall it calls it Dere Street and B6309. Also, official addresses in Whittonstall only give the address as "The Cottage, Whittonstall", with no street given, so that's no help.

Anyway, I think the current values are as accurate as we can be, but please to edit if you find better sources.

I've had a look at other sources and relabelled the route from Whittonstall to Ebchester. Dere Street is the name of the Roman Road so only applies to the parts that intersect with the road. Chare Bank is the official name in Ebchester, but I can't see if it extends beyond the bridge.

An entirely valid point and I wasn't aware that I'd added the name (though I clearly did!) as I entirely agree with you. The name is on the Roman road relation now, so I've removed them. Thanks for spotting.

It was from https://www.achurchnearyou.com/swainsthorpe-st-peter/
but I agree it's a long way off (it seems to be the p/c for Great Melton church). I'll remove it. Thanks for spotting.

Nice spot. Looks like my copy/paste hadn't cleared from the previous one I did. Corrected now.

Over-exuberant copying of OS_OpenStreetView, thanks for spotting. A quick web search shows it did close in 2012. Now reverted.

you're right -- I have gotten over-excited there. I've reverted it, thanks.

Thanks -- I was there at the weekend and was hugely impressed by the variety of shops. It certainly deserved immortalising on OSM!