[OSM-talk] Improving ref=* documentation

[Greg Troxel]

"Brian M. Sperlongano" <zelonewolf at gmail.com> writes:

> If the sign on the ground doesn't match the government's database, then the
> obvious answer is that the government database is wrong.  I don't see why
> we would replicate demonstrably wrong data into OSM.

That doesn't follow at all.  It could be that the signs are wrong.

This is often tricky, but when a government authority has the legal
right to name something, their records of naming decisions are the
primary source.  That some other government employees made a signing
mistake is another issue, but it doesn't invalidate the legal status of
the authority to name.

And, these naming decisions are entirely verifiable; you just go to town
hall and ask to look at the records.   Which is basically the same thing
as looking at a sign, just a bit more time consuming.

I've seen this around me, with minor variations from what I think are
the official names (which would be in the minutes of the Selectboard),
MassGIS maps, and signs.  It's on my todo list to rationalize it all and
file what are effectively bug reports.  But it's things like "Deer Field
Lane" vs "Deerfield Lane" that are not significantly confusing to
humans.


If the database is a map database that is intended to store a copy of
what the sign is, and the sign is what matters, then you are right.  But
without figuring out which messy case we're dealing with, I don't think
we can sign "signs are right" for road names/refs.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20210804/b4cb7f6a/attachment.sig>