ravenPrivacy
What this website records.
What we record.
When you open a page on this site, the page itself sends us a short note. It contains:
- The page you opened — its path, like
/download/. Never the query string, so anything you or anyone else appended to the link is dropped before it is stored. - Where you came from — only the host, like
news.ycombinator.com. Not the page on it, and nothing at all when you arrived from another page on this site or typed the address yourself. - The campaign tags in the link, if it had any — the
utm_sourceand friends we put on our own posts to find out which ones worked. - Reading time and scroll depth — how many seconds the tab was actually in front of you, and how far down the page got. A tab you opened and forgot about over lunch counts the seconds you looked, not the hour.
- Two coarse words about your setup — one of macOS, Linux, Windows, iOS, Android or other, and one of phone, tablet or desktop. Six values and three values. Not a fingerprint, and not your browser’s full user-agent string.
- A click on a download link — which published file, and its version. The files themselves come straight from our CDN, so this is the only reason we know a download was ever started. It is a click; we cannot tell whether the file finished.
That is the complete list. There is nothing else in the table.
How you are counted, and why it is not you.
To tell “one person read two pages” from “two people read one page each”, a site needs some way to say same reader. The usual way is a cookie, which is a permanent marker left in your browser. We do not do that. Nothing is written to your browser’s storage except one random number that identifies the tab you are reading in, and disappears when you close it.
Instead, our server takes your IP address, your browser’s user-agent, today’s date and a secret only we hold, and hashes all four together into a short string. Your address is used for that calculation and then thrown away — it is never written down. The hash cannot be reversed into an address, and because the date is part of it, the whole scheme changes at midnight: today’s string cannot be matched to yesterday’s, by us or by anyone who ever saw the table.
What this buys is one narrow fact — two pageviews, one reader, today — and it deliberately buys nothing else. We have no way to recognize a returning visitor, no way to follow anyone between pages over time, and no identity to connect any of this to a Raven account.
What we don’t do.
- No third parties. Every script on this site is served from this site. No analytics product, no advertising network, no social widget, no session recorder. Nobody else receives anything about your visit, because nobody else is involved in it.
- No cookies, and no consent banner. The second follows from the first: we do not ask you to click through a dialog, because there is nothing to ask permission for.
- No selling or sharing. The numbers exist so we can decide what to write and what to build. They are not a product, and there is nothing in them anyone could buy.
- No profiles. We do not build a record of a person, because nothing we store identifies one.
Where it lives, and for how long.
The rows go into our own database, on our own infrastructure, alongside the rest of the Raven service. They are kept for 400 days — long enough to compare one release to the one before it — and then deleted automatically.
A named handful of people on our team can read the totals. What they see is this page’s list added up: how many people read which page, for how long, and which downloads were started. There is no view of an individual reader, because there is no individual in the data to view.
If you would rather not be counted.
Turn on Do Not Track or Global Privacy Control in your browser and this site sends nothing — not a pageview, not a second of reading time, not a download click. No setting of ours to find, no preference to remember, and the site works exactly the same.
A content blocker will also stop it, and we have no interest in working around one.
The forms.
The feedback form is different, because you are choosing to tell us something. What you type reaches our team along with the email address you optionally include, and we use that address to reply to you about your report and nothing else. Leave it blank and we simply have no way to write back.
The request-access form is the one place on this site that keeps what you wrote. Your name, work email, the tier you asked about and what you told us you would use it for are stored in our own database — the same one the service runs on, no form product and no third party — because a request is a conversation we owe you an answer to. We use it to reply and to decide what to build next, and nothing else. Ask us to delete it and we will.
The app and Raven Teams.
This page is about this website. Raven itself, and the Raven Teams service you sign in to, are separate systems with their own accounts and their own data — a signed-in service necessarily knows things a public webpage does not. Ask us at feedback@thousandbirds.ai and we will tell you exactly what they hold.
Check us on it.
None of the above needs to be taken on trust. The code that sends these measurements is one file on this site, the code that stores them is one file in the service, and the table they land in is one migration — all three in the same repository as the app you downloaded. A claim on a privacy page is worth what the code behind it says, so if something here does not match what you find, that is a bug and we would like to hear about it.