How Safari Pretended to Be Mozilla Before It Was Released

Former Apple employee Don Melton is sharing a unique look behind the scenes of the Safari development team. Melton was the team leader on both the Safari and WebKit products that are now used by millions of users on both iOS, the Mac, and Windows.

Previously, Melton explained how the Safari name came about, but today he shares the tale of Safari's User Agent string and the strategies his team used to keep the project under wraps.

Twitter and Facebook didn’t exist then. Nobody at Apple was stupid enough to blog about work, so what was I worried about?

Server logs. They scared the hell out of me.

When a Web browser fetches a page from a Web server, the browser identifies itself to that server with a user agent string — basically its name, version, platform, etc. The browser also gives the server an IP address so the server knows where to return the page. This exchange not only makes the Web work, it also allows the server to tell who is using what browser and where they’re using it.

You can see where this is going, right? But wait, there’s more…

Back around 1990, some forward-thinking IT person secured for Apple an entire Class A network of IP addresses. That’s right, Apple has 16,777,216 static IP addresses. And because all of these addresses belong together — in what’s now called a “/8 block” — every one of them starts with the same number. In Apple’s case, the number is 17.

IP address 17.149.160.49? That’s Apple. 17.1.2.3? Yes, Apple. 17.18.19.20? Also, Apple. 17.253.254.255? Apple, dammit!

I was so screwed.

Melton's blog has the rest of the details about how his team kept things quiet before the big reveal.