[Imports-us] A proposal for handling the tiger realignment

Serge Wroclawski emacsen at gmail.com
Sat Aug 10 02:36:28 UTC 2013


Eric can (and should) elaborate, but the bottom line is that the
algorithm used is not "Has the way's geometry been modified", which
has led us to this problem.

Other issues I see with an automated import:

1. Even with the TIGER re-alignment, we'll need another import process
for roads that are not in OSM
2. We'll also want something for roads which were in TIGER, but are no
longer in TIGER (eg due to construction) or roads which have either
been renamed, or been corrected in TIGER.
3. Even in the best situation, there's detail that TIGER sometimes
omits, such as turning circles.

On the other hand, users are not ideal either:

1. Properly training OSM users is a non-trivial task (and I don't
think it's a good idea for non-OSMers to be involved in imports)
2. Even OSMers need extra training for use of Josm for the task of importing
3. Users make errors, get lazy, or miss details, especially as the
data increases in size

The hope with my proposal is to remove as many barriers to getting
improved data for OSM, but also try to involve the community.

It's a really delicate balance.

- Serge

More information about the Imports-us mailing list