Privacy

One thing, and only if you change it:

• Your light or dark preference, in localStorage under rpcnode-theme. It never leaves your browser and is never sent to us. Clearing site data removes it.

No cookie is set for reading the site. The only cookie this platform issues is a session cookie for the operator console, which requires an account that only the people running this deployment have. If you have not logged in, you do not have one.

The same thing every web server sees when you request a page: your IP address, the URL, the time, the response status, your user agent and — where your browser sends one — the referring page.

Two uses, both operational:

• Rate limiting. The public API is limited per address, so a counter keyed on your IP exists for a short window. It is held in memory, expires on its own, and is not linked to anything else.
• Diagnosing errors and abuse. Request logs are kept short-term to find faults and to identify traffic that is degrading the service for other readers.

These logs are not analysed for behaviour, not joined to anything, not shared, and not used to build a picture of any reader. They are operational exhaust, kept because a server that logs nothing cannot be debugged.

There is a third use, and it only happens if you ask for it: on Nearest to me an IP address is turned into an approximate location so that endpoints can be suggested for it. That address is normally one you type in rather than your own, nothing about it is stored, and it is set out in the next section.

One page, Nearest to me, answers “which endpoint suits where my code runs”. To do that it turns an IP address into an approximate location and finds the closest place we operate a probe.

Four things are true of it, and they are the reason it is a section of its own:

• The address is yours to give. The page asks for the public address of the host that will make the calls — usually a server, and usually not the machine you are reading this on. Choosing a region instead skips address handling altogether, and no address is read or sent on that path.
• Your own address is only ever a suggestion. The address your connection appears to come from is shown beside the field so you can use it if it happens to be the right one. It is not submitted for you and no result is computed from it unless you ask.
• Nobody else is involved. The lookup runs on our own server against a database file we host, the same one used for the node map. No geolocation service is contacted, so no third party learns the address or that it was asked about.
• Nothing is kept. The address is not written to a log, a table or a cache. The location derived from it exists for the length of one request and is discarded with it. Responses from this page carryno-storeso that no cache between us holds an answer computed for you.

The result is an approximation twice over, and the page says so rather than burying it: a database places an address to a city at best, and the figures shown were measured from our nearest probe rather than from your host. Where that probe is thousands of kilometres away, the page states the distance and says the answer describes our coverage more than your location.

Two places. Both are disclosed here because a privacy page that claims “no third parties” while the browser is talking to one is worse than no page at all.

• Map tiles. The network map loads its basemap from an open tile service, which necessarily sees your IP address and which tiles you asked for. Every other page on this site loads nothing from anywhere but here. If you do not want that request made, do not open the map.
• Links you click. We link out to block explorers, provider documentation and specifications. Following one is a normal visit to that site under their terms. Our outbound links carry noreferrer, so the destination is not told which of our pages you came from.

There is no font CDN, no script CDN, no embedded video, no comment system and no social widget. Everything else the page needs is served from this origin.

Most of what this site publishes is machine data: response times, error classes, block heights, node addresses. It describes infrastructure rather than individuals.

Two categories deserve saying out loud, because they look like personal data and are not:

• Validator identities are public keys published by the chain itself. We index what the chain publishes, and we attach no identity to them beyond the name their operator chose to publish.
• Node and validator locations are the approximate geographic location of an IP address, resolved from a public database. They describe where a machine is routed — almost always a datacentre — and are never a claim about where a person is. Addresses that do not resolve stay unresolved rather than being assigned a country.

If an address published on this site is a home connection rather than infrastructure and you would like it removed, write to us and we will remove it.

Where data protection law applies to you — including the UK GDPR and the EU GDPR — you have the right to ask what we hold about you, to have it corrected, to have it erased, and to object to its processing.

In practice the answer to the first is usually “request logs from the last few days, which we cannot tie to you without your IP address and the approximate time”. We would still rather you asked than assumed.

Where we process anything that is personal data, the lawful basis is legitimate interests: running a public service, keeping it available, and preventing abuse. We have kept what is collected to the minimum that permits, which is why there is no analytics.

privacy@rpcnode.io — access, correction and erasure requests

If this changes — if analytics is ever added, or a third-party service introduced — the date at the top of this page moves and the change is described here rather than quietly made. Adding a tracker without saying so would contradict the entire premise of the site.