I am curious.
Considering “Don’t map historic events and historic features” what is the taught about this:
The 1838 version has a start_date and end_date, the 1840 version has start_date, end_date and date, and the 1901 version has no date related keys at all.
It looks like that was intended for Openhistoricalmap.org but got accidentally entered into OSM.
You mean the original creation of these boundaries some years ago, or the most recent changeset that changed police_district to police? It looks like the latter was aimed mainly at current boundaries in Hungary, with a possibly accidental side effect for historic boundaries in Ireland.
Looking at the age of the data, no. This relation was added in 2016 with an explicit OOC source. The original OHM server (and project) fell apart around 2015/2016, so there wouldn’t have been an OHM server to add the data to at that point.
Something was restored in 2017 but missing much of the original data.
There absolutely is an argument that something that was valid between 1838 and 1840 would make more sense in OHM now, but that would need:
Previous discussions with OHM people with regard to (1) have only resulted in “the person who added the data to OSM is free to move it if they wish”. See this other thread (among others). Personally, I’d have thought that the people behind OHM would be keen to see well-sourced and dated information such as the relation here in OHM, but there has been no offer of assistance to do that.
With regard to (2) I suspect that includes at least this site. I don’t know about others.
One final caveat with regard to the Irish situation - the mapping of “old” boundaries was in part inspired by the legal status that they still have today (a point someone made in one of the Irish chat channels a couple of days ago), so it’s not a valid assumption to say that “something with a start date of 1838 obviously has no relevance in the present day”; you’d need to check with the local community to make sure. Even if it does, it might still make more sense to be in OHM (given that the current iteration of that supports start and end dates explicitly), but that will require effort from both the people behind OHM and current data consumers - it won’t happen by magic.
Outside of the general case, with regard to these specific boundaries, I’d suggest discussing the tagging with the person that added them - either via a changeset discussion or via a topic in the Irish forum here. If you do the latter you’ll want to let the person who added the data know about it, since I don’t think that they are active on this forum yet.
Ah, thanks, I didn’t notice that these were older boundaries that were just recently added. The start and end tags jumped out at me because that’s what OHM uses.
I do have to quibble a bit with:
Why is there an “us” and “them” here? OSM is the free map that anyone can edit. OHM is also the free map that anyone can edit. As an added bonus, OHM even uses the same toolset1 as OSM. Whenever someone wanders into the OSM forum saying “hey you OSM folks have got things all jacked up” for one reason or another, the usual response is, “you do know you can edit the map yourself, right?” Same with OHM.
I do like your suggestion about making things easier for data consumers though. It seems like it would be a pretty straightforward to make the OHM planet file accessible under the Community vector tile server without too much effort, which would make historic features easily available for anyone sitting on an mvt vector stack.
Indeed, OHM is the wiki world map that anyone can edit. 
@SomeoneElse might have been frustrated about what I said in that thread about copying from OSM to OHM. It’s not that OHM’s Powers That Be don’t want to help OSM mappers; it’s that we want to be a good partner to the OSM community, and a big part of that is respecting OSM’s license. Otherwise, what message does this send to people who would co-opt OSM for commercial gain?
Anyways, there was a similar discussion recently about these historic census boundaries and the relationship to townlands:
As you know, the U.S. community is more cautious about importing census boundaries. But the arguments in favor of the Irish census boundaries are slightly different, that they have legal force and are useful for genealogical research. This makes them quite analogous to survey boundaries, which the U.S. community has a complicated relationship with, especially in New England.
The easy answer is that all these boundaries certainly belong in OHM, but as long as there’s also some attachment to them in OSM, that critical mass undercuts any local interest in building out OHM’s coverage. Without that interest, OHM would be taking, not partnering. Rather than offering to take the Irish community’s hard work, OHM’s offer first of all is to seek a partnership with the community and help them get their feet wet in the project.
Anyone is welcome to use the first-party OHM vector tileset for their projects without restriction (though credit to OHM would be much appreciated). The schema isn’t stable and there are lots of known issues, but the development team is actively working through them. The tileset updates minutely at higher zoom levels and slightly less frequently at lower zoom levels.
An alternative tileset would be cool too, of course. The existing OSMUS tile service is somewhat geared towards the OpenMapTiles schema, which isn’t quite ready for OHM data (no dates), but Planetiler is capable of many things that are challenging with the older Tegola-based stack.
It was partly that - but to be clear here the source of the original data is clear, so it’s not just “copying from OSM”. I’m not a lawyer, but don’t believe that:
would be a violation of OSM’s licence.
Obviously, many of the same people active in OHM are also active in OSM, but the important group here is “people who just want to see their historical data on a map and don’t care in which project it resides”. There are plenty of examples where it’d be beneficial to both projects if OHM were a little bit more helpful towards both its potential data providers and data consumers. Former railways, (e.g. under new housing estates) is one; but there are others.
I changed one word above (OHM->OSM), and suddenly the statement takes on a very different light, I think.
I think people that complain about OSM not being helpful could make a choice to be part of OSM and be the change they want to see.
The same is true of OHM.
I think you’ve missed the point entirely.
OSM has numerous resources that the community has made that help potential data providers and consumers to (to borrow a phrase) “switch to OSM”. OHM does not have that - or if it does, I’m not aware of them, and if I’m not aware of them, I’d suggest that a casual user of OSM data for historic purposes would also not be aware.
Those resources were written by people who cared about the things that they were writing about. Broadly speaking, the Irish community hasn’t objected to things like Dublin Metropolitan Police District 1838 being in OSM, and if they don’t mind, I personally don’t mind either. I can absolutely see the logic for a move, but requires someone who cares about the data to move it. That person might be someone who added the data, someone who uses it now, or someone who cares about the OHM project (likely it’ll be a combination of all three). It isn’t me.
The “Reuse” page on the OSM Wiki is our “Switch to OHM” guide for prospective data consumers. You can find a link to it towards the top of the OHM wiki portal. Please spread the word! If you’d like to contribute your know-how toward a standalone site similar to Switch2OSM, feel free to crib from that page, and ping the forum if you run into any questions translating from OSM to OHM.
As for data providers, @jeffmeyer and I have practically worn out our welcome on the Digital Humanities symposium circuit pleading with academics to share their data with OHM, and there are lots of existing resources for tracing old maps and other OHM-specific workflows. However, you have a point about needing to document more. The wiki doesn’t have a lot about how to import data into OHM. I think that’s primarily because everyone who’s done imports so far has just winged it, often without any discussion. For better or worse, pretty much everyone who contributes to the project so far has had to roll up their sleeves and blaze their own path.
You might recall OSM being like this 15 or so years ago, before Process began inserting itself into community norms. Process benefits those who aren’t as bold in breaking new ground, but comes from experience and a bit of scar tissue. We would eventually like to have Process too, but first we would need more Experiments, both successful and otherwise, that we can tout as case studies.
In the meantime, OHM’s usual process goes like this when anyone catches wind of a promising data source:
In time, someone who’s interested in this topic will pick it up based on these starting points. The community is around to provide moral, if not technical, support. This could be less passive, but we’re doing what we can with the resources we have at the moment. If you know of anyone who can help to champion this sort of stuff from the inside, we’d welcome their help every bit as much as the actual importing.