#19520 closed enhancement (wontfix)
Uploading incomplete mulitpolygon relations - rather block than warn?
Reported by: | Owned by: | team | |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Core | Version: | |
Keywords: | Cc: |
Description
I propose that the validator should block upload and not offer the "Continue" option, when it prompts "Incomplete multipolygon-relation modified" (may not be the exact wording). Perhaps offer a "Download elements" option instead, or, good enough, select the relation(s), so one does not have to search for it.
Rationale: Its "warning" only, but may hide an "error". Recently I destroyed a large multipolygon because I did somehow create an intersection in it. Trying to find out, how that could have happened, I guess it is because I moved a node, close to the border, maybe even outside of the downloaded data area, and later click through the warning - Blocking should not unreasonably burden mappers much as it is quick most of the time.
Attachments (0)
Change History (5)
comment:1 by , 5 years ago
comment:2 by , 5 years ago
Hmm, there must be more at play - just right now I pulled a node of an inner area "shingle" outside of its outer ring in a "water:river" multipoly; Hitting the upload button, no error, no warning. Everything completely downloaded. Am I confused?
comment:3 by , 5 years ago
I just opened #19526 - as can be seen from case 2 there, the validator needs all the members present to find certain errors in multipolygon relations. (Also, the validator needs to run on the relation, first ;)
I mostly work like this: Open a GPX and a number of geotagged photos, then walk the trace and "Download in current view" as I go along. So I come over a number of relations, and as long as I do not touch any of them, no problem. Multipolygons tend to be large and winding and I too rather not have all their members in view from the start, as to stay focused. But I would not mind, if, before uploading I was forced to fetch them, for the validator be get it right. A broken MP, as I have witnessed, can damage the map quite a lot.
comment:4 by , 5 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Looks like 16761/josm fixes this in a better way. Closed by opener.
comment:5 by , 5 years ago
Resolution: | fixed → wontfix |
---|
No, please, not as long as the test is not improved. See #18610.
Why is this strict warning only raised for modification of MPs. Split way, does break PTv2 relations under certain conditions, atm, but I do not get warnings about modified route relations.