I think that it is a good summary (disclaimer - I am doing it an extremely tiny scale, with extremely small dataset of points).
Exact implementation details depends on what you exactly need - detect deletion in OSM? detect missing data in OSM dataset/foreign dataset? Detect differences that will appear? Produce dataset containing both OSM data and data from foreign dataset (may be necessary if for some reason that data is not importable into OSM and it is OK to produce ODBL dataset). Jul 21, 2020, 06:02 by ahhug...@gmail.com: > Hi All, > > We expect to encounter the same problem at the NHVR if we begin to use OSM. > > My (possibly unfounded) initial thoughts are based around linking the OSM & > Source feature outside OSM in something similar to a "join" table. The join > might be on attribution (id), geometry or both. Then, you have to accept that > the link/join will break and a process is needed to detect breakages when > they happen so they can be repaired (a mix of automated & manual). > > Someone else might be able to comment on this with more clarity. > > The way I see it, you can't stop the breakage. You have to accept it and deal > with change. > > A Hughes > > > On Sat, 18 Jul 2020 at 23:10, Sebastian Spiess <> mapp...@consebt.de> > wrote: > >> On 9/7/20 7:52 pm, Mateusz Konieczny via Talk-au wrote: >> >>> >>> >>> >>> Jul 9, 2020, 06:50 by >>> greg.dutkow...@gmail.com>>> : >>> >>>> Hi, >>>> Bicycle Network Tasmania are trying to improve the quality of >>>> cycling infrastructure information in OSM. >>>> Much has been done by volunteers in various jurisdictions, and >>>> we have done lots locally, but the tagging is quite complex for >>>> cycle paths and not always correct. >>>> Local councils are responsible for much of the infrastructure, >>>> but they usually have little interaction with OSM. >>>> It would be most efficient if the councils GIS data worked in >>>> tandem with OSM data so that they kept each other up to date, >>>> each storing the info that is most useful for them. For >>>> instance, for bike parking, there is little utility in OSM >>>> storing the asset numbers and other info that the councils use >>>> to maintain their assets (although the ref tag could be used as >>>> a foreign key to help keep the two in sych). >>>> The Hobart councils we work with are concerned with the quality >>>> of the data in OSM and the ability of anyone to change it. >>>> Does anyone know of any examples we could learn from of local >>>> government itself working to keep OSM data up to date? >>>> Thanks. >>>> >>> One of the easiest things that local government may do is to >>> >>> 1) publish their datasets on an open license allowing to use it by >>> mappers >>> 2) react to reports of mistakes in their data >>> >>> Both work relatively well in Poland for address data - with >>> publishing required by >>> national law (though still ignored be many local governments) >>> >>> Note that (1) is useful for mappers even if data quality is >>> unsufficient to import it >>> into OSM. I am using a bit noisy bicycle parking in locating >>> unmapped ones >>> (often location, description and real location mismatches >>> significantly, but >>> almost always it allows me to find something that was missing in OSM) >>> >>> _______________________________________________Talk-au mailing list>>> >>> Talk-au@openstreetmap.org>>> >>> https://lists.openstreetmap.org/listinfo/talk-au >>> >> >> Hi, indeed great to see you reach out. >> >> >> Yes I agree that a good approach is to make the data open. However, I >> understand Greg is asking if there are working concepts on how to >> maintain a link between local government GIS (which might have >> additional information) and OSM data. >> >> >> Once the relevant information has been entered into OSM, how is the >> council to track the data? e.g. to see if tags get modified, nodes >> moved, added. >> >> >> e.g. worst case is that a nicely mapped and tagged area gets re-done by >> someone. This results in new node and way numbers. >> >> >> A good example would be a single node gets expanded by OSM users. >> >> >> In both cases the data is diverging from another. How to keep track? >> Are there concepts/solutions? >> >> >> Yes >> >> _______________________________________________ >> Talk-au mailing list >> >> Talk-au@openstreetmap.org >> >> https://lists.openstreetmap.org/listinfo/talk-au >>
_______________________________________________ Talk-au mailing list Talk-au@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-au
