Facebook scales. It’s not easy to scale computing infrastructure to support billions of daily active users. Only a handful of companies can do that right now. On top of that they scaled an ad business that can target billions of people. There is only one other company in the world who can do that.
> What technology has Facebook developed with all of those "best people" that could compare to what Bell Labs did with a group of fewer than 10?
See it’s less about “quality”, it’s more about “quantity”. Scaling is a different problem on its own. I adore Unix but I admire what Facebook did as well. Scaling any project over 10 exceptional people is one thing and scaling a project with 10,000 engineers is a completely different thing and Facebook succeeded at that.
Whether that’s a good thing or not that they have succeeded is a separate discussion.
I don’t doubt that delivering an online service, even a crappy service, at FB scale presents some unique and difficult challenges, and I’m sure FB employs lots of smart and talented people. But given the number of supposedly “best” engineers they do employ the results seem underwhelming. I find it more impressive that Amazon manages to keep AWS up and reliable while constantly introducing new services, and that Apple manages to support a billion iPhone users without spewing ads in their face.
While the process was a trainwreck that I'd never entertain again, I will admit that while on campus, I met and talked to some really, really smart people.
I just remember being kinda sad that all this talent was being wasted on ads and holding attention.
Back to the point, I think anyone who has been there could attest it's not all just leetcode studiers...well, maybe some amount, but they employ a lot of really smart people.
Maybe best at grinding LeetCode, best at playing office politics, best at tolerating toxic work environments...
But likely not best in actually building good software. No doubt quite good at that, but probably not meaningfully different than any other tech company in the area hiring from a similar talent pool.
And because the candidates also need to be optimized for the qualities mentioned in the first paragraph, likely there’s a trade off where average software development skill ends up being lower due to needing to maximize other personality factors first.
They clearly have engineers that are good, but he questions if the engineers are better than the good engineers found at any other run-of-the-mill company. They aren't exactly performing sorcery over there, just solving problems like every other software business.
That doesn't mean it's any less impressive of the people who did build all this, but I just don't think we need to compare, we can recognize the accomplishment on its own.
But that's not what we're talking about. The question posed was if Facebook's good developers are better than good developers at run-of-the-mill companies. That is not clear. People at Facebook, with an average tenure of only 2.3 years, have almost certainly also worked at run-of-the-mill companies so that already suggests that you can find good developers all over the place.
That developers at Facebook solved problems Facebook had doesn't prove no other developer could also solve Facebook's problems if they also worked at Facebook. Solving problems is literally the job of a developer, so it seems highly unlikely that all other developers outside of Facebook cannot also solve problems. Of course developers are only going to solve the problems in front of them.
Due to the absurd amount of SWEs on Facebook, surely someone or a team happen to working there will score a huge FOSS hit sooner or later just due to the numbers.
Not to mention creating some of the widest-used OSS libraries on the web.
Plus, it’s one of the few platforms our grandparents can use, and we all know how hard it can be to make software that older folks like to use.
So they are clearly not hiring terrible devs. (Well, sure some of them are, but that’s the case everywhere)
My parents successfully use iPhones and iPads and GMail, along with Facebook. I can’t say they “like to use” Facebook, but if they want to see photos of their kids and grandkids they have to use it. I don’t know anyone who enjoys using FB, including my own kids. It seems users see it as a necessary evil — where else can they go?
Any metric you like. Facebook is massive and has a huge range of people. There'll be people who sit in the top 0.1% for literally every method of ranking them you can dream up.