[Talk-us] Multipolygonizing

[OSM Volunteer stevea]

Kevin and others:  please do not misunderstand me.  There ARE times when shared ways between multipolygons is an elegant and THE correct solution, as you, I and many others have found to be true and edited into existence many times.  By no means do I advocate that where such beauty has been completed that it be torn apart and reverted to simple polygons, that would be a giant step backwards.  These solutions are NOT "too complicated," and it is a misunderstanding of what I have been saying in this thread to think so.

Your statement that "mechanical edits (often by the JOSM tool reltoolbox) running roughshod over carefully curated data" strikes at the bullseye of what I wish to convey.  Such trampling really makes updating imported (curated) data quite difficult, and OSM really DOES want to encourage the updating of imported data.  I AM saying:  please BE CAREFUL multipolygonizing polygons, especially where tagging or changeset data might indicate they are part of an import or a curated set of data.

It may be that as other geodata, especially those which align well with the idea that "shared ways are a good idea in these data," become better aligned with OSM's multipolygon data structure, this situation greatly improves.  Now, there is some alignment, though it is not perfect; for example, shapefile data imported into JOSM are either "OK after import" or "come close enough to easily fix" (in my opinion).  But concomitant with this is that OSM editors — software, novices, intermediates and experts alike — not be afraid of or intimidated by relations and/or multipolygons (and editing them).  While our "primitive types" of nodes and ways are relatively easy to learn, relations are not, but we MUST prioritize it as an important task that even beginning users better familiarize themselves and gain comfort with these more complex types of data — early, and often.

Please, ENTER data using shared ways where it makes sense to do so.  Nobody is saying "don't do that."  ALSO, please be aware that existing NON-multipolygon data (especially imports and other "curated" data) may very well suffer from the process of being "multipolygonized."  There is a balance to be struck, and I would be very disheartened to see our map become "dumbed down" by data which should be multipolygon somehow become twisted into not.

SteveA
California