Joe Longstreet

An inventor / maker living in Kansas City, Missouri


realistic human placeholder photos

Whenever I drop in placeholder images, I like them to be as realistic as possible. If the image is a placeholder for a user profile photo, it’s important to use a poorly lit, crazily framed, pixelated, weird looking person. Why? It gives you a better sense of the design decisions you’ll need to make to deemphasize user uploaded content.

Introducing Faceholder, a tool for designers and developers to utilize realistic photos of people.

I find quite a bit of utility in this app and use it frequently. What was surprising is how interested I am in all of these people I don’t know. I found myself pulling up Faceholder and just looking at pictures of random people, so I added a links to their profile. Now you can just browse and click people forever.

How it works

In the background, the app is repetitively doing a procedure. An abbreviated itinerary:

  1. Pick a random number between 1 and a billionish (roughly the population of Facebook).
  2. Ask Facebook for the profile URL of a user with an id that matches the random number.
  3. Bounce around to a bunch of different URLs until Facebook finally hands off the profile image of the requested user.
  4. Save a reference to it.
  5. Repeat FOREVER.

The first few times I ran the above procedure, I found 90+% of the photos collected were exactly the same. The image was the default photo for users who hadn’t yet uploaded profile pictures. Strange…, I wonder what percentage of Facebook users are robots or dead accounts and what percent are actual human users.

A billionish “users” * 10% (a very very generous 10 percent) of human photos ≈ 100 million real humans on facebook.

I have no idea if the above is correct, but thought it was interesting. If you're interested, checkout the github repo.