Good question.. Based on my experience with Nominatim I really, really hope the answer would be that it imports faster/uses less space (to be clear: I have no idea if it does - but imports and the sheer data volume have been my biggest pain point with Nominatim)
What inspired you to start this project? There are numerous open source geocoders and the core problem isn't the software anyway, it's data. I'm often perplexed by people spending time building new things thinking they've uncovered/solved something when researching/talking with others who have domain experience would quickly reveal there may not be a there there.
I need to solve a few problems for my city:
1. Search crossroads
2. Search pair of lat/lon from non structured text
3. Possibility to add a custom data
And it was my inspiration to create yet another geocoder.
Love the itch scratching motive! Crossroads and intersections is a nice idea - they are used quite a lot in American cities, from casual observation "at the corner of Pine and Union".
I imagine in some countries with different addressing systems, road intersections may be a primary means of locating things.
14 comments
[ 6.4 ms ] story [ 40.8 ms ] thread[1] http://wiki.openstreetmap.org/wiki/Nominatim
EDIT: Codebases appears to be entirely different though (Node.js vs Go)
And it was my inspiration to create yet another geocoder.
I imagine in some countries with different addressing systems, road intersections may be a primary means of locating things.
Not to forget OpenStreetMap and https://mapzen.com/blog/who-s-on-first/