[fn1]

Off-topic, but how the Facebook, Messenger or Instagram-app need a million lines of code is absolutely beyond me.

I had a few encounters with facebook's opensource code (mostly around react-native) and it is generally subpar in my opinion.

[aeyes]

Facebook is primarily a message and photo sharing application but it also has group boards, dating, live video streaming, a marketplace, browser games and probably more features. Then there are all the features for targeted ads, tracking, payment and so on. All of this is available almost all around the world.

How do you implement these features for more than a billion active users in less than a million lines of code? Even 100 million sounds like a low estimate.

Why does the Linux kernel have more than 30 million lines of code?

You could ask if these features make sense. But in the end it is their business decision to make.

[gpderetta]

how does the number of active users affect the size of the application?

[mkl95]

My guess is that it's an extreme case of Conway's law, with many teams that produce mountains of code hidden behind some API. Not to mention codebases provided by external firms, which could amount to a large % of all those KLOC.

I'm interested in how they avoid "Roman empire syndrome", as in owning more software than they can maintain while actively expanding it.

[barbariangrunge]

Rome held on to its land for a while. More like napoleon or Alexander syndrome

[tantalor]

This happens with "everything" apps that have hundreds of features you have never seen, each serving a tiny fraction of the userbase who somehow stumbles on them and justifying the PM/eng resources to keep them chugging along, or abandoned and left to rot until these migrations come along.