[Talk-us] NHD preconverted data - posted

[Richard Fairhurst]

Ben Supnik wrote:
> I can definitely do that - I never got vector backgrounds working with
> Potlatch myself (I have a Trac issue floating around) but if you can
> make Potlatch read the hosted files directly, that'd be a nice win.

Vector background layers in P2.2 are _much_ improved over earlier
versions. I've just downloaded one of your NHD .zips, extracted a file
from it, hosted it locally
(http://127.0.0.1/~richard/osm/01010003_nhdwbhi0.xml, that sort of thing),
and imported it into P2, and it's very happy with it. Have a play if you
can put an unzipped file somewhere.

> Besides one .osm per zip, what else does Potlatch need?  For example,
> how does Potlatch need the data to be indexed/named?  Right now the data
> is named by HUC (the entiy formerly known as sub-basin) + data layer, so
> unless you know what kind of data you want and the USGS's number for
> your town, you can't really easily guess the filename.  There's
> something like 2300+ HUCs.
>
> I can build an index off the particular shapefile HUC boundaries if
> that's needed.

I think an index is the way to go - certainly that's what I'm encouraging
people to do with the parallel case of OS VectorMap District data in
Britain. If you could put a trivial Nominatim gateway on the front of it,
that would mean someone could type in a town name like, oooh, "Potlatch,
Idaho", and be presented with the direct .zip links, ready to paste into
P2.

cheers
Richard