Hi, Fredrik:
Thank you for the effort you put in preparing this import.
I've had a look to the wiki and the data, and I don't see major problems
so that this import can be safely done, but I've taken 8 random church
nodes (don't have much time to get a bigger set) and I've seen some
issues that you should take into consideration to improve the import
workflow, before it starts.
---
1) -42734 (Mæl kirke):
"architect"="Linstow type" -> Shouldn't be Hans Linstow [1] instead?
That church (that is already mapped) is inside an area tagged as
amenity=graveyard. Maybe some instructions to correct that, would make
the import better (at least with some examples of what users involved in
the import may find during the task).
2) -42440 (Valjok kirke):
This church already mapped has two name tags in OSM:
"name"="Valjok kirke"
"name:se"="Váljohka girkko" -> name in Sami language
In this case, would you add the Sami name to the name tag?
In general, what will be the sintax when you add both names to the name
tag, so all users use a consistent naming? Will it be "Valjok kirke -
Váljohka girkko", "Valjok kirke / Váljohka girkko", "Valjok
kirke;Váljohka girkko"...?
3) -39128 (Terningmoen garnisonskapell):
No issues detected.
4) -42800 (Bø kyrkje):
"name"="Bø kyrkje" -> But the wiki isn't clear on what you will do when
it is already mapped with a different name. In this case, the name in
OSM is "By kyrkje", so which one will you keep if names are different?
The church is also mapped as an area (building footprint). But the wiki
doesn't clarify what footprint will remain, if the one present in OSM or
the new one from the import. In case this last one replaces the former,
it would be very advisable to keep its history (using CTRL+SHIFT+g of
JOSM's UtilsPlugin2 plugin).
Also, if you substitute the old footprint by the new one, you have to
take care of the grass area [2] attached (in OSM) to the church
footprint. Team involved in the import should know how to solve these
situations.
5) -40194 (Andebu kirke):
"architect"="ukjent" -> ukjent means unknown, isn't it? I guess in this
case it would be better to skip the architect tag, or change it to
"architect"="unknown", although it seems there isn't any object in OSM
with the "architect"="unknown" key&value combination according to
taginfo [3]. I've checked, and there are 266 churches with this
"architect"="ukjent" tag.
6) -41110 (Stavang kyrkje):
"architect"="Ole Halvorsen, Bergen" -> It seems Bergen means the
architect is from Bergen, but looks not part of his name. In that case
it shouldn't be there.
7) -42044 (Nordvik kirke):
No issues found.
8) -41866 (Gløshaug kirke):
A building footprint is already present. Same question as with church #4
(Bø kyrkje).
---
Browsing around all tags, I see some issues:
end_date=* is generally discuraged. And you shouldn't use it for all 7
churches that have this tag. If a church was a church and is not a
church anymore (it is a community centre, for example), it can't be
building=church. Let's say the church was constructed in 1780 and ended
as church in 1978, and now is a community centre. In that case it should
be "building:1780-1978"="church",
"amenity:1780-1978"="place_of_worship", "name:1780-1978"="Whatever
Kirke", ..., "building"="civic", "amenity"="community_centre",
"name"="Whatever Neighbours Friendship Centre", etc. Please check the
lyfe cycle wiki page for more info [4].
ruins taggins is discouraged in some cases. In this dataset we have 10
churches tagged as "historic"="ruins" + "ruins"="church". It's better
"historic"="church" + "ruins"="yes". See the ruins wiki [5] and the life
cycle wiki [6].
In the wiki it says "The 1951 data points will be imported and merged
with existing churches with the importers editor of choice.". But
towards the end, in the Workflow and Conflation sections, you say that
users will use JOSM. I guess it's safe if you just ask the team involved
to use JOSM only, so you ease the workflow and tools/plugins to use.
At least the architect tag values should be revieved, and set some rules
for the import team to make the corrections consistent.
Cheers,
Rafael.
[1] https://en.wikipedia.org/wiki/Hans_Linstow
[2] https://www.openstreetmap.org/way/493988037
[3] https://taginfo.openstreetmap.org/tags/?key=architect&value=unknown
[4]
https://wiki.openstreetmap.org/wiki/Comparison_of_life_cycle_concepts#Date_namespace_suffix_as_.22keyname:DATESPEC.3D...22
[5] https://wiki.openstreetmap.org/wiki/Key:ruins
[6] https://wiki.openstreetmap.org/wiki/Comparison_of_life_cycle_concepts
On 27/01/18 17:10, Fredrik wrote:
Evening,
An import has been planned and set up for importing all 1951 churches in
Norway from official sources.
The import is described here:
https://wiki.openstreetmap.org/wiki/Import/Catalogue/Kirkebyggdatabasen
and the Norwegian mailing list has been contacted and discussed.
The data is available from links on the import page as well as all
details concerning the import
--
FredrikLindseth
openstreetmap.org/user/FredrikLindseth
_______________________________________________
Imports mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/imports
_______________________________________________
Imports mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/imports
