6 ms·

I'm a co-founder at WonderProxy, we didn't make their list (we target people doing application testing, not consumer VPNs). We're in 100+ countries, and I'll s

I'm a co-founder at WonderProxy, we didn't make their list (we target people doing application testing, not consumer VPNs). We're in 100+ countries, and I'll stand by that claim. It's a huge pain in the neck. In our early years we had a lot of problems with suppliers claiming to be in Mexico or South America who were actually just in Texas. I almost flew to Peru with a rackmount server in my luggage after weeks of problems, that plan died when we realized I'd need to figure out how to pay Peruvian income tax on the money I made in country before I could leave. We've also had customers complaining that a given competitor had a country we'd had trouble sourcing in the Middle East. A little digging on our part and it's less than a ms away from our server in Germany.

reincoder 9mo ago
I work for IPinfo. I have raised a ticket internally, but I think we focused on consumer VPNs for this test. For our ProbeNet, we are attempting to reach 150 countries (by ISO 3166's definition). We are at around 530 cities. Server management is not an easy task. We do not ship hardware, but operate using dedicated servers, so this reduces one layer of complexity. To maintain the authenticity of our server locations, we utilize cross-pings and network traffic behavior detection. If any abnormality is detected, the server will be immediately disabled to prevent polluting our data. There will be a ticket to investigate what went wrong. We pay for each (excluding 3 to 4 servers where the owner and the team really likes us and insists on sponsoring) server. Expansion is an active effort for us, as there are 70k ASNs and about 100 more countries where we do not have a server. We hope to partner with more ASNs, particularly residential ISPs and IXPs. So, a lot of effort is put into active outreach through WhatsApp, emails, social media and phone calls. We use a number of different data-based techniques to identify "leads".
- laz 9mo ago
  Google, Apple, and Meta (maybe others?) have the data to build a complete GeoIP dataset. None of them will share because there are only downsides to doing so. When FB was rolling out ipv6 in 2012, well meaning engineers proposed releasing a v6 only GeoIP db (at the time, the public dbs were shit). Not surprisingly, it was shot down.
  - reincoder 9mo ago
    We are always happy to work with large technology enterprises and streaming platforms, not necessarily to sell, but to share insights, data, and practical advice. We observe the entire internet through active measurements, and we are open to co-publishing research when it benefits the broader ecosystem. Google/GCP is top of mind for me due to a recent engineering ticket. Some of our own infrastructure is hosted on GCP, and Google’s device-based IP geolocation model causes issues for internet users, particularly for IPv6 services. From what we understand, when a large number of users from a censored country use a specific VPN provider, Google's device-based signals can bias the geolocation of entire IP ranges toward that country. This has direct consequences for accessibility to GCP-hosted services. We have seen cases where providers with German-based data centers were suddenly geolocated to a random country with strict internet censorship policies, purely due to device-based inference rather than network reality. Our focus is firmly on the geolocation of exit-node IPs, backed by network evidence. https://community.ipinfo.io/t/getting-403-forbidden-when-accessing-the-ipinfo-api-and-website/303 https://community.ipinfo.io/t/getting-403-forbidden-when-acc... We are actively looking to connect with someone at Google/GCP, Azure/Microsoft and others who would be willing to speak with us, or directly with our founder. Our community consistently asks us to partner more deeply with enterprises because we are in constant contact with end users and network operators. To be honest, we do not even get many questions or issues. We are partners with a large CDN company, and I get one message about a month, which usually involves sharing evidence data and not fixing something. From a large-scale organization's perspective, IP geolocation should not be treated as an internal project. It is a service. Delivering it properly requires the full range of engineering, sales, support, and personnel available around the clock to engage with users, evaluate evidence, and continuously incorporate feedback.
    - ACCount37 9mo ago
      > From what we understand, when a large number of users from a censored country use a specific VPN provider, Google's device-based signals can bias the geolocation of entire IP ranges toward that country. Yep, this is a known effect. How it seems to work is: Google uses Android phones as data harvesting probes. And when it sees that a lot of devices in a given IP range pick up on GPS data, Wi-Fi APs or cell tower IDs that are known to be located in Iran, and possibly other cues like ping to client devices or client device languages, timezones, search request contents, then the system infers "there's a network wormhole there with Iran on the other end", and the entire IP range grows legs and drifts towards Iran. The owner of those IP addresses can mitigate the issue, mostly by shaping traffic or doing things to Google's system, but I know of no way for anyone else to do it.
- YouAreWRONGtoo 9mo ago
  [dead]
- Youden 9mo ago
  Could you use RIPE Atlas and its network of probes, at least to fill in areas where it's difficult to get your own probes? That way everyone benefits.
  - reincoder 9mo ago
    We are actually a sponsor of RIPE Atlas and have a bunch of credits. But I am not sure if we use them extensively. I think, as we own and operate the ProbeNet, much of the data collection efforts can be done through that in a scalable manner.
- preinheimer 9mo ago
  We really don't want to operate our own hardware. The situation in Peru at the time was that there wasn't anyone offering the bandwidth we needed who could actually back up their bandwidth claims. Forget 95th percentile, bandwidth there was straight "you pay for a pipe, we give you that size pipe (but somewhat oversold)". But no one could do more than like 5mbit that was actually more like 3.

reincoder 9mo ago

I work for IPinfo. I have raised a ticket internally, but I think we focused on consumer VPNs for this test. For our ProbeNet, we are attempting to reach 150 countries (by ISO 3166's definition). We are at around 530 cities. Server management is not an easy task. We do not ship hardware, but operate using dedicated servers, so this reduces one layer of complexity. To maintain the authenticity of our server locations, we utilize cross-pings and network traffic behavior detection. If any abnormality is detected, the server will be immediately disabled to prevent polluting our data. There will be a ticket to investigate what went wrong. We pay for each (excluding 3 to 4 servers where the owner and the team really likes us and insists on sponsoring) server. Expansion is an active effort for us, as there are 70k ASNs and about 100 more countries where we do not have a server. We hope to partner with more ASNs, particularly residential ISPs and IXPs. So, a lot of effort is put into active outreach through WhatsApp, emails, social media and phone calls. We use a number of different data-based techniques to identify "leads".

laz 9mo ago
Google, Apple, and Meta (maybe others?) have the data to build a complete GeoIP dataset. None of them will share because there are only downsides to doing so. When FB was rolling out ipv6 in 2012, well meaning engineers proposed releasing a v6 only GeoIP db (at the time, the public dbs were shit). Not surprisingly, it was shot down.
- reincoder 9mo ago
  We are always happy to work with large technology enterprises and streaming platforms, not necessarily to sell, but to share insights, data, and practical advice. We observe the entire internet through active measurements, and we are open to co-publishing research when it benefits the broader ecosystem. Google/GCP is top of mind for me due to a recent engineering ticket. Some of our own infrastructure is hosted on GCP, and Google’s device-based IP geolocation model causes issues for internet users, particularly for IPv6 services. From what we understand, when a large number of users from a censored country use a specific VPN provider, Google's device-based signals can bias the geolocation of entire IP ranges toward that country. This has direct consequences for accessibility to GCP-hosted services. We have seen cases where providers with German-based data centers were suddenly geolocated to a random country with strict internet censorship policies, purely due to device-based inference rather than network reality. Our focus is firmly on the geolocation of exit-node IPs, backed by network evidence. https://community.ipinfo.io/t/getting-403-forbidden-when-accessing-the-ipinfo-api-and-website/303 https://community.ipinfo.io/t/getting-403-forbidden-when-acc... We are actively looking to connect with someone at Google/GCP, Azure/Microsoft and others who would be willing to speak with us, or directly with our founder. Our community consistently asks us to partner more deeply with enterprises because we are in constant contact with end users and network operators. To be honest, we do not even get many questions or issues. We are partners with a large CDN company, and I get one message about a month, which usually involves sharing evidence data and not fixing something. From a large-scale organization's perspective, IP geolocation should not be treated as an internal project. It is a service. Delivering it properly requires the full range of engineering, sales, support, and personnel available around the clock to engage with users, evaluate evidence, and continuously incorporate feedback.
  - ACCount37 9mo ago
    > From what we understand, when a large number of users from a censored country use a specific VPN provider, Google's device-based signals can bias the geolocation of entire IP ranges toward that country. Yep, this is a known effect. How it seems to work is: Google uses Android phones as data harvesting probes. And when it sees that a lot of devices in a given IP range pick up on GPS data, Wi-Fi APs or cell tower IDs that are known to be located in Iran, and possibly other cues like ping to client devices or client device languages, timezones, search request contents, then the system infers "there's a network wormhole there with Iran on the other end", and the entire IP range grows legs and drifts towards Iran. The owner of those IP addresses can mitigate the issue, mostly by shaping traffic or doing things to Google's system, but I know of no way for anyone else to do it.
YouAreWRONGtoo 9mo ago
[dead]
Youden 9mo ago
Could you use RIPE Atlas and its network of probes, at least to fill in areas where it's difficult to get your own probes? That way everyone benefits.
- reincoder 9mo ago
  We are actually a sponsor of RIPE Atlas and have a bunch of credits. But I am not sure if we use them extensively. I think, as we own and operate the ProbeNet, much of the data collection efforts can be done through that in a scalable manner.
preinheimer 9mo ago
We really don't want to operate our own hardware. The situation in Peru at the time was that there wasn't anyone offering the bandwidth we needed who could actually back up their bandwidth claims. Forget 95th percentile, bandwidth there was straight "you pay for a pipe, we give you that size pipe (but somewhat oversold)". But no one could do more than like 5mbit that was actually more like 3.