Spiders from Osmose

[brawer]

Osmose has some spiders written in Python, mostly for data feeds in France. However, other than AllThePlaces, Osmose doesn’t just fetch the data and convert it into a common format; it also matches the data against OSM and proposes data edits. Of course, this also means that an Osmose “merge analyzer” is more complicated, less self-contained, and (for newbies) more difficult to write than an AllThePlaces spider — I’m not at all suggesting that ATP should take this approach. But it might be worth porting some or all of Osmose’s existing spiders into AllThePlaces. https://github.com/osm-fr/osmose-backend/tree/master/analysers

[CloCkWeRX]

This is probably worth looking at in detail, comparing licences, etc.

@matkoniecz is more or less building the same ATP to OSM conflation tooling as described above. Potentially the better approach here is to expose the data from that process in the same kinds of places osmose data gets integrated?

[matkoniecz]

I may try, though I had relatively poor experience with Osmose and their approach to validity/quality of listed issues. Maybe I have overly high expectations about quality of what QA/Validator/edit suggester should report.

From what I see at

https://github.com/osm-fr/osmose-backend/issues?q=is%3Aissue+is%3Aopen+ATP

https://github.com/osm-fr/osmose-backend/issues?q=is%3Aissue+is%3Aopen+AllThePlaces

https://github.com/osm-fr/osmose-backend/issues?q=is%3Aissue+is%3Aopen+All+The+Places

they were not considering use of ATP data so far.

[matkoniecz]

I opened https://codeberg.org/matkoniecz/list_how_openstreetmap_can_be_improved_with_alltheplaces_data/issues/7 to track this

[matkoniecz]

And got reply indicating that it would not be welcome in Osmose.

Pulling some data sources used by Osmose into ATP, as suggested in this ticket still may be a good idea.

[matkoniecz]

@brawer Which Osmose reports use data that seems to be usable also by ATP and not used yet?

[brawer]

I don't have a list, we'd need to check each one.