Hacker News new | ask | show | jobs
by ggm 394 days ago
They said they are scraping other sources. So it's the RIR data from whois, rdap or the delegated files, modulated through curated lists like maxmind, and collections from GEOFEED sources as defined in the rfc.

There are other models. I'd be surprised if they do any triangulation (for instance) because it's success rate in finding errors or misreporting in the above is low, unless you are Akamai and have 400+ points of BGP to get closest match from.