Threads

Profile picture of @binhonglee on Twitter. @binhonglee

First of all, reported for misinformation. Alright, so this tweet bothered me enough (especially looking at other people eating this up at face value) that I guess I want to rant about it so here's a thread on why this doesn't mean anything.

Profile picture of @NehemiahKiv on Twitter. @NehemiahKiv

The bigger the product the more lines of code used?

Read more...

1. I highly doubt you can get an accurate LoC count on any of these apps considering that they are not open sourced and they evolve constantly (even on a day-to-day basis).

2. Even if we assume it's accurate, what's the methodology here? Like do you count all the GCP and FB's in house infra code? Are you gonna add all of AWS count to Snap's LoC count as well since they use AWS? What about tools? Like UI for monitoring graphs or whatever?

3. What about programming languages? FB wrote Hack and have it in the same repository as like everything else. (This is how the language API can evolve quickly.) Are you going to count LoC of like C++, Java, Kotlin, ObjC, Swift etc. into a lot of these apps too?

4. Kinda similar to 3 but frameworks. FB has all of React as part of the code base and Twitter uses React for their web UI. Are you gonna add LoCs for React to Twitter's as well?

5. How do you meaningfully separate IG / WA code count from FB? And YouTube from G? They might all live in separate repositories / folders but there's a good chance they reuse a lot of the libraries and services. Are you just going to double / triple count them?

I think I wrote enough at this point. Well that's it for now. Please don't fall for these sort of clickbait images. Thanks.

10/23/2021