I've been using Robert Whittaker's resources to try to correct postcode errors
in OSM. After discussion with another mapper, I'd like to have the opinion of
the community as to what we should put in OSM. A simple answer might be to show
the correct, current postcode (i.e. if the Postcode Address File became
open-source we'd import that). However there are a couple of things to think
about:
As mappers we use the information that is on the ground. For postcodes that
could mean till receipts, company websites, FHRS or edubase data. What if these
do not show the current correct postcode and there is no open source which
shows the correct one? Is an incorrect postcode better than none? What if one
of those sources (say FHRS) shows the correct postcode but more obvious sources
(e.g. the company website) shows an old postcode? I can use the correct
postcode but it is likely to be changed back to a wrong one by a mapper who has
only found the obvious source.
I've noticed that many bank branches have a unique postcode while the shops
either side of them all share a single postcode. When the bank branch closes
(as so many do these days) the unique postcode finds its way on to Robert's old
postal district list. A new occupant of the bank premises is likely to be given
the same postcode as the other shops nearby. Do we need to retain the old
postcode for any historical purpose? If so we probably also need to tag that
this building was once a bank but if we do that for every previous occupant of
every shop things would soon get out of hand! While the bank is empty and,
hence, no longer listed on the PAF, is it better to retain the old postcode or
delete it and show no postcode for that building?
I want to improve the map but don't want to delete or overwrite anything which
may be of some value. Thoughts please!
_______________________________________________
Talk-GB mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/talk-gb
