Ian Dees

We pull data from authoritative address providers and don't modify the data on its way through our pipeline. To update the data, we'll need to find a newer dataset to...

Thanks for the suggestion! We already do this, but currently only show the latest downloads on the project page. You can see the history of builds (along with download links)...

The archives go back farther (at least a couple years), but I was doing it on another AWS account and I haven't updated the index JSON files to point to...

Maybe if we ask really nicely...

I was thinking either, but was mostly joking. We can try using Cloudflare's "friendly bots" form mentioned [here](https://blog.cloudflare.com/friendly-bots/)?

I don't really want to adopt OSM's multiple-tag system for specifying categories (amenity=restaurant, cuisine=burger means "burger restaurant"). I think this is pretty difficult for data consumers to understand. I do...

Another requirement for categorization would be to make it relatively easy for spider authors to find acceptable category options. Ideally the categories we pick would have an easy list to...

Name Suggestion Index IDs would be good to add, but I don't know that they would help non-OSM data consumers. That's probably OK, because those data consumers could look at...

@thekuiperbelt @wessport do y'all have any thoughts on categories? Do you apply categorization when you consume the data?

> I know alltheplaces has a [Gap scraper](https://github.com/alltheplaces/alltheplaces/blob/master/locations/spiders/gap.py) - I guess I'd have to look more closely at what it actually generates. I don't know if this happens now, but...