Remove codemeta

FIX #409

mention github as primary site in readme, and list other locations

Remove env vars to skip tests

sf is at G2139. We are at the last version 😎https://earth-info.nga.mil/php/download.php?file=WGS%2084(G2296).pdf

All tests run in 10s

Avoid skipping test-unname.R for test-coverage action

Implement osmdata_sf(..., out = "meta")

The final Z in the timestamps indicate UTC timezone
(https://en.wikipedia.org/wiki/ISO_8601#Coordinated_Universal_Time_(UTC)),
but as.POSIXct() doesn't take it into account and
should be specified with parameter tz = "UTC".

FIX #269

Implement described viewbox parameter in getbb()

Wasn't functional

getbb() can query wikidata ids for OSM relations properties (P402)

Idea from @mhpob
FIX #401

rm sourcehut remote

fix sourcehut remote url

try 'github.ref' instead of '.head_ref' to trigger workflow

update syntax for 'github.head_ref'

remove initial workflow condition entirely

reinstate 'push' condition in push-to-elsewhere workflow

@jmaspons the 'on: push: branches: main' is actually needed, because
a PR merge triggers the push event, and without that, this doesn't
run at all

update remote urls for #388

getbb() with empty results

FIX: #392

That's just too clunky, and will be unable to cleanly describe
how to set up remote mirroring there. All others are much easier.

Use functions from R base instead of methods

@mpadge: There are some conditions under which x <- x@polygons does not
give the same result as x <- slot (x, "polygons"), and that the latter is
a more stable and thus preferable.

Drop methods from imports

from ?methods::is
"Although inherits is defined for S3 classes, it has been
modified so that the result returned is nearly always equivalent to is,
both for S4 and non-S4 objects. Since it is implemented in C, it is
somewhat faster."

fix elevation to close #389

FIX #377 and improve getbb() for polygons and sf_polygons()

Same result either with one matching place_name or more