Two changesets from the same user making the same mistakes, deleting highway=trunk tags from some ways, which obviously makes them non routeable.
I’m not an iD user, but, it is easy in iD to make such mistakes? the user is warned of what is doing?
Two changesets from the same user making the same mistakes, deleting highway=trunk tags from some ways, which obviously makes them non routeable.
I’m not an iD user, but, it is easy in iD to make such mistakes? the user is warned of what is doing?
You probably meant Way: BR-471 (776990721) | OpenStreetMap and Way: BR-392;BR-471 (476315093) | OpenStreetMap.
If you select “highway” in the tag part, yes you can do it. Or change the type to “line”.
And iD won’t warn you.
However, as you’ve seen other edits from this contributor have been reverted.
It’s rare - people in iD tend not to do this. That said, If I do edit the area and delete the highway tag (like here on the dev server) no warning was displayed.
I guess that alternative would be to strengthen ban on tagless ways and consider attributes such as name/ref to be insufficient.
Not sure is there an issue proposing this.
No, some are useful, if they’re member of a relation, a multipolygon for instance. Except that, a warning makes sense.
I have reviewed hundreds of changesets by newbies (iD users) for the last 13 years. One can delete almost any tag if they really want to. I do not think that it has been a problem. Vandalism can happen, it belongs to OSM because we are open.
When newbies contribute how can a software distinguish between good-faith edits and bad-faith edits? Do we want to limit our online editor to a online version of StreetComplete. Please keep in mind that we have highly (almost daily) active contributors using iD as their main or only editor.
My initial concern was about changes that broke routing on trunk or primary roads. The question was whether there was something effective yet non-intrusive to mitigate that problem. I’m not framing this as vandalism, but rather as a mistake by a novice or distracted user. I’m not proposing this for all elements in OSM, but particularly for these roads where, if the error is not detected by some mechanism -either before/during editing or afterwards through appropriate quality controls- it would generate problems in one of the most important features of OSM, and these would not be local problems limited to a small area, but potentially affecting larger areas given that we are talking about trunk roads.