Talking with Herman Martinus about analytics on Bear, I found myself wanted to store only one bit of data: the HTTP referer.

I don’t care how many people have visited by blog, but I care where they came from.

Thus, my ideal analytics would probably look something like this:

Unique referrers would be stored and displayed alphabetically.

By default, Bear Blog uses CSS to trigger analytics with the following bit of code:

body:hover {
    border-image: url("/hit//?ref=");
}

All I need to figure out is, what’s the easiest possible way to store this data? I don’t need to make it private, I really just need the barest of databases and HTML to make it look presentable.

Herman Martinus got back with an easy enough solution:

Yeah, I reckon that’d be easy enough to do. You can set up an endpoint that just logs every referrer, then pop something like this in your blog’s footer:

if (document.referrer) {  

    const url = new URL("https://example.com/log-referrer");  

    url.searchParams.append("referrer", document.referrer);  

    fetch(url);  

}