normally in OSM we say that we map stuff that is visible with one’s own eyes and can be verified “on the ground” - and we then make some (limited!) exceptions when we’re talking about data that is important enough to warrant an exception, like administrative boundaries.
I have accidentally clicked the query button on a random location in Dublin and have been informed that it lies within the following areas:
a couple different administrative areas, from Ireland down to Oldbawn, with a few oddities like a “Vice county” and a “Barony” thrown in;
29 different electoral boundaries (e.g. “Dublin Constituency” for 6 different years, 8 polygons all called “Electoral Boundary Dublin South West”, “Tallagth South” for 5 different years)
18 further different “historic” boundaries, e.g. “Historic Site Dublin South Superintendent Registrars District 1885” or “Historic Site Tallaght No. Three DED 1971”
three different statistical regions
a tourist region
a “Poor Law Union”
a “Dublin Metropolitan District Court” as well as a “District court area Rathfarnham District Court Area”
and some more
Most of these are not surveyable, and must have been imported from third-party data sources. In many countries the importing of such an amount of historic data would be met with skepticism by the community. A cursory inspection shows that many (but far from all) of the boundaries above have been created by user @brianh, and as expected, many are in the high double digits of version numbers because anyone unsuspectingly splitting a way that forms a part of such a boundary will have edited the boundary.
Is there a consensus in the Irish community that “you can import any historic boundaries of any kind as long as you have some sort of document as a source” (a random pick lists the source “GRO pamphlet 1885 HMSO in NLI Ir 9141 B 35”)? Or have these imports been discussed and agreed upon individually?
Because according to “what is usual elsewhere”, we’d probably get rid of at least 80% of the above relations and ask people to add them to OpenHistoricalMap instead.
Many of these points were discussed in this earlier thread (the thread title refers to a specific location but the discussion quickly moved to a general conversation about historic boundary mapping in Ireland):
Essentially, it’s well known that Ireland is an outlier in this respect. From time to time people pop up to say “wouldn’t these boundaries be better in OHM?” As far as I know nobody has ever popped up to say: “I am willing to do the work to add these to OHM and rework any tools that currently use these boundaries in OSM, and here is my plan to do it.”
I’d also note that many current boundaries in Ireland were not available under a suitable license and had to be traced from historic out
-of-copyright maps or constructed from text descriptions, so it’s not surprising that the general idea of tracing from old maps became well established, and it’s not a big jump from there to recording historic boundaries from similar sources.
(Full disclosure: I am from Ireland but don’t live there and don’t do much mapping there. I’m answering mainly because I think relatively few Irish mappers participate in this forum).
As I understand it, those are still somewhat relevant.
Those, obviously not.
Absolutely that. Because of the “unique way that OHM is constituted”, there isn’t a “lift and shift” option. No-one from OHM has said what people who use this data could do to move it, or offered to help beyond vague “you can repeat your work if you want”.
Without explicit complaints from Ireland (which I’ve not seen) I don’t think that there is a pressing need to move it. It is a shame that OHM doesn’t seem to want it - it does seem right up their alley.
Edit: Private conversations suggest that there is more of a conversation ongoing than public discussions would suggest. Until that discussion is more public I suspect we’ll keep seeing topics like this one.
Most historical boundaries in Ireland can still be seen on the ground, though. Pretty much the entire island is divided into farmers’ fields. The boundaries are hedges, streams, lanes & roads… the vast majority of them are all still there, and have been for centuries.
Yes. Modern boundaries like electoral constituencies often run along pre-existing roads. Legacy (mostly pre-1900) roads were often built along pre-existing boundaries.
Yes, there is more going on than one would know by following the OSM forum. Fortunately, I don’t think copyright will be a major impediment to transferring the historic boundaries to OHM. Our current obstacle is more mundane: teaching mappers to do some gymnastics in JOSM so only the relevant data transfers over cleanly. We all know how difficult it can be to juggle all the Irish boundaries that have built up in OSM. We want to make sure it’s done well the first time. Quite honestly it’s a bit of “deaf leading the blind” – really shouldn’t be an iD-head like me showing people around JOSM!
Please understand that everything in OHM moves at a glacial pace compared to OSM. OHM has a much smaller community but a much larger scope. I myself am like a dozen messages behind on this front. Calling individuals out probably won’t make them more comfortable discussing their efforts in public. If anyone would like reassurance that some progress is being made, they can follow changes in OHM publicly just like one would follow changes in OSM.
Beyond the boundaries, there is one simple thing that would really help for strengthening cooperation between the two projects. If any mapper is OK with their original contributions (in whole or in part) being in OHM in the public domain, they could say so on their OSM user profile. It would help us more effectively find willing participants and low-hanging fruit.
Does that waiver already exist somewere so I can copy it onto my profile as well? I already searched in this forum, the OHM forum and the wiki, but I couldn’t find anything.
There isn’t a standard code or anything like that yet. But someone could add a link to their OHM profile to their social links, then add to their profile description something to the effect of “I release my original contributions [of X/in Y/since Z] into the public domain.” Then at least an OHM mapper would have more confidence that this user’s edits wouldn’t obstruct transferring an individual element over. This may sound very similar to the public domain declaration, but unfortunately the public doesn’t have access to that declaration.