It’s been almost a year since my first day at Facebook, sitting in an overcrowded meeting room with my bootcamp class because 42 Earlham Street was full and it’d be another week before we moved to 10 Brock Street, with its gargantuan empty spaces (which are no longer empty: nearly half the company has joined since I started).
How has that year been, you ask? Hard. It’s been hard. Fun, worthwhile, educational, exciting, but hard. Facebook is very different from any company I’ve worked at before: it’s bigger, it’s faster, it’s more ambitious. Everything I had learned about making software outside the company was…well no, not wrong, certainly not wrong. But it was perhaps inapplicable. Facebook does things differently, I can’t immediately have impact by doing what I did at my other employers, how can I cope with this?
When you’re used to the life of minor engineering celebrity, where people go to conferences because you’re on the bill, and your pithy statements on programming get retweeted up the wazoo, it’s easy to get hubristic. I did: I know what I’m doing, the engineers here don’t yet know that, and there will be a glorious revelation when I bring the two tablets (an iPad and a Galaxy Note 10, I don’t know) down from the mountain, sell everyone in the company a copy of my book and we all start doing things The Right Way™.
That wasn’t going to work. Not because what I wanted to do couldn’t work, but because I was starting from the wrong place. How many billion-dollar companies had I written software for when I started at Facebook? None. So who was I to say that my way of writing software was “right”? All I could say was that it definitely felt right.
What I had to learn to move past this was that Facebook works on data. If I can show that there’s time being lost, or bugs being introduced, or run times being lengthened, and that what I want to do would save time, or catch bugs, or speed things up, then Facebook will listen.
The opposite of “my way is not incontrovertibly correct” is not “my way is incontrovertibly incorrect”, but it was easy to come to that conclusion too, and to rage-quit the idea of having any effect here. I think I avoided this, but I also think I came pretty close to it. I could easily have decided that while I don’t like Facebook’s way, I need to suck it up and work that way. That’s wrong for the same reason that trying to boil Facebook’s oceans was wrong: just as what I’m used to doing isn’t necessarily correct, what Facebook is doing isn’t necessarily correct. There’s room for change.
Indeed, Facebook engineering is far from change-averse. Everything is set up to let you make changes, from the extreme collective code ownership (“bored changing the code you normally work on? Change the rest of it!”) to incident review (“OK, that change wasn’t the best, let’s see what we can learn from it”).
So, it took me a long time, perhaps longer than it should: I used my first year to realise that I can change Facebook and Facebook can change me, and the way we’ll make this happen is by showing each other how the changes will make us better. Now, as David Bowie said, it’s time to turn and face the strange.
