automatically pull OSM data for validated feeds

[nickolasclarke]

placeholder

[nickolasclarke]

@r-barnes your input on best source for this would be valuable.

[r-barnes]

One way to manage this would be to download an OSM world file (50GB), use an OSM parser to extract just the road networks (dropping it to a few GBs), and then building a parser that looks at the extent of the GTFS data, crops out a generously-margined OSM extract, and builds the contraction hierarchy from it.

Alternatively, we could try to download all the separate files from interline and link them with the feeds. I think that would be a largely manual process, but haven't thought about it extensively.

[nickolasclarke]

I was thinking about the former option as well. Its a heavy initial lift, but well within grasp. 50 GB is very manageable to store, and this strikes me as the simplest solution, esp given it would not need to be updated with great frequency.