[OSM-talk] Was the deletion of Null Island reasonable?

[stevea]

On Jan 2, 2022, at 7:35 PM, Michael Collinson <mike at ayeltd.biz> wrote:
> 
> I don't have a rigid opinion though I do feel that if someone wants to know where something is then there is a didactic (instructive/educational) case, strong or not, for putting it in OpenStreetMap.
> 
> Null Island does not physically exist but then neither do boundaries and it could be considered a point boundary and very precisely located (as distinct from the Bermuda Triangle). On the other hand, tagging it as an island is certainly not appropriate, 'cos it ain't. So on balance, the buoy, the name, wikipedia and a description tag seem reasonable. Adding place=locality might also meet consensus; I don' see much difference between it and the hundreds of names dotted across the rural moorland areas of my homeland that cannot be associated with a precise physical feature.

+1 to all of this.  I offer my nod of approval to a place=locality tag, that comes pretty close to if not exactly describes what this is.

Honestly (though, I haven't done any "empirical studies"), Null Island (as a well-documented node) might exist solely as a place which prevents bad data entry, for a certain combination of savvy editors (who know what Null Island "is") and those who bumble their way into having this or a corner of it inadvertently creep into our data, but are "saved" by this node being there and informing "this is an odd little 'corner' of Earth, with nothing actually there."  So maybe you want to check your edit.  C'mon, that's really happened.  I know I would smile a little bit and back out my mistake, blow away some eraser crumbs, be glad that GOOD data in our map "informed me" (the potential error of my ways) and all ends well.