I understand your frustration, but in general, and in some specific cases, active measurement at our scale (1300+ servers, self-operated, 550+ cities) is going to be better than geofeed. Again, we appreciate geofeed and we accept geofeed to be important. But we are approaching data granularity for some cases that have never been seen or even imagined to be possible in the IP geolocation industry, even though this industry is nearing three decades old.
Please understand that a portion of network operators being attentive and vigilant about geofeed is not universal. We definitely appreciate them, but they are a positive outlier. We absolutely respect them for their vigilance and contribution. For us, we operate on a global scale and our data impacts everyone. Bigger telecoms do not maintain geofeed accuracy. Outside of active and conscious participants a lot of time we do not see geofeed's existence, let alone maintenance of it. Our product is not specific to a handful of willing and geneorus network operators; it is specific to all IP addresses for which we have to provide data. Geofeed's data verification is such a significant issue; we have focused on active measurements, which have resulted in universal gross accuracy improvement for our data. The infrastructure, the data, the team, and the mission require massive investments. For a hosting IP address, we cannot point to the rack or room, but we can at least point to a data center because we have a PoP hosted in the data center. With active measurement, we can also point the neighboring IP addresses to the data center. In general terms, active measurements the way IPinfo does (not the industry as a whole) is a superior method to the current status quo. And when active measurement has gaps, we can always refer to geofeed. We are not ignoring geofeed; we extend and enhance it to a degree. When geofeeds are used to misdirect, we do ignore it. It goes back to three things: Verification, accuracy, and granularity. For now, active measurements are the reasonable path that makes every internet user happy. We are not representative of the industry. My opinions are about IPinfo's approach to IP geolocation. We are trying to carve out a separate path for because of the challenges we have seen in the industry. We are trying to be present in every discussion with network operators because we do not want frustration to be the reason for collaboration. We genuinely are trying our best for network operators to be happy with us. If your data is wrong with us, tell us, reach out to us, we will work out a solution. But our current approach to IP geolocation has in general been largely satisfactory in practice. — Abdullah | DevRel, IPinfo Help us improve our data: https://forms.gle/Ja64QWrJuW8PkUzK6 _______________________________________________ NANOG mailing list https://lists.nanog.org/archives/list/[email protected]/message/A7EL6KC3W6KT63AV5S4S3DUTXWBGMXUN/
