Upon further analysis of empty relations, I suspect they will be far more problematic than I expected. While it is on the wiki since 2010 and feels like a powerful tool, it does not seem to be used (let alone supported by consumers).
My bad, I should not recommend things that I have not used myself. I am curious as to the recommended approach - we have a similar parking zone system here in Prague and I have no idea how to map it. A relation still sounds like a horrible solution due to the parking depending on the side of the road and due to the sheer frequency of road splits due to (for example) added surface tags. Maybe a wikidata tag? On 14 January 2018 at 15:05, Matej Lieskovský <lieskovsky.ma...@gmail.com> wrote: > Citation provided: > https://wiki.openstreetmap.org/wiki/Relation#Size > https://wiki.openstreetmap.org/wiki/Relation:route#Size > > Notice that the border relation you linked is already version 790 (and > borders change far less often than roads). Viewing the relation on osm.org > already takes some time on my PC, I'd hate to have to import it into JOSM. > If you don't think that is bad enough, try working with this relation: > https://osm.org/relation/912311 > > On 14 January 2018 at 14:45, Mateusz Konieczny <matkoni...@gmail.com> > wrote: > >> On Sun, 14 Jan 2018 12:32:02 +0100 >> Matej Lieskovský <lieskovsky.ma...@gmail.com> wrote: >> >> > if you group all the streets in a single relation, the relation is >> > likely to be rather big. >> > This can be hard on the server. >> >> [citatation needed] >> >> We have massive boundary relation >> ( https://www.openstreetmap.org/relation/49715 for example) and AFAIK >> everything works fine. >> >> Is this performance concern is confirmed by anything? Messages from >> people handling OSM servers, profiling, anything like that? >> > >
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
