It's rare to find someone who worked for them that's dissatisfied with the experience. They employ nowhere near the amount of people that FAANG employs, but people that work there are usually really passionate about their work and receive good benefits in return.

Point being that the author is more of a rule than an exception.

Disclaimer: Was a Mozilla fellow a few years back.

Mozilla was both an amazing and a highly frustrating company to work for, when I was there from 2010 to 2015.

It was also a very different place when I started (before Fx4) from when I ended (just before FxOS was killed).

When I started it was a place full of passion, with a lot of technical vision going on (Fx4 was a major reboot and there were a number of side projects going on that showed promise), albeit not necessarily a lot of obvious strategic vision. I'm sure there was more behind the scenes with John Lilly, whose leadership I hired into, but I lost confidence after he left and suddenly it seemed like the message was "desktop is dying, mobile is everything."

Wasn't our mission success based on having enough market share to win a seat at the standards tables and win a place on the "supported browsers" test plan for major websites? Getting a significant part of the shrinking desktop market we'd already executed well on in the past and that competitors were idling on might be better than getting a little of the mobile market that companies with greater resources were bouncing off of left and right, no?

Intranets and SaaS apps are still a thing, and offices still use desktop, so there'd still be a core audience right? Maybe mobile browsers can be different and less standardized than desktop browsers and that's OK? Maybe it has to be? Maybe it even should be while mobile browsing incubates? Maybe browsers won't even be the primary way websites interact on mobile and they'll use client apps instead?

That was a confusing pivot for me at the time, and Mozilla's strategy was to both put all the momentum on mobile and to kneejerk to a rapid update model for the desktop browser, inspired by Chrome. Problem was that destroyed the desktop add-ons community because it turns out you can't do that when you have a monkeypatch/binary extension model with high coupling, and Jetpack/Add-Ons SDK wasn't mature enough or powerful enough yet to replicate most existing add-ons.

It also exhausted the users because the existing flow of having to explicitly approve updates on launch still remained, only now it was frequent enough to disrupt workflows--you never knew when launching a web page meant having to navigate the updates dialog first. Google had designed their browser ecosystem and usability around that update model, it wasn't something you could just graft on. By the time we figured it out Chrome had picked up a decent chunk of the community.

The company then more or less doubled in size, in no small part bringing in a bunch of people from mobile and related sectors that didn't have the FOSS culture in their backgrounds. That culminated in the development of FxOS, which I always felt like was treated as an unwelcome fork by the platform team. Maybe it was because of the need to support two fundamentally different forms of interaction, two different models of security, two different distribution and update models, two different lots of things in wrapping Gecko with an application vs. wrapping it with an OS. That also divided the company, since there were now two broadly different technical missions going on, albeit sharing code.

When considering the success of FxOS vs. KaiOS, it's worth thinking about the drag having two competing priorities in the same company causes, and how that might clarify when the 3rd party is doing the fork instead. Conflicts like "how do you release a fix for Fx the browser when it'd zero-day FxOS the phone and you can't get an update through the carrier for two months" may not be so much of a problem to figure out without that tension. I'm extremely impressed with what Fabrice and co. have able to pull off with KaiOS, and I bet lightening that load helped a lot.

When I left, after it was plain FxOS was not going to succeed at that time, in that environment, it was still a place full of passion--but it was now also a place equally full of frustration, and not with a lot of strong leadership going on and a sharply muted FOSS culture. Seemed like there were a lot of missteps and platitudes, but not a ton of optimism. I was very happy to see Servo come to fruition and Quantum be released, because I honestly expected Mozilla to go down in flames before they could get the desktop browser to a state where people who'd defected would install it again.

I'll always treasure my experience with Mozilla--and having worked at two FAANGs now since, I agree the level of talent was equal or greater at Mozilla. But I do have to admit I wish I'd shifted that five year period about 2-3 years earlier across the board. I came in right after the really good part, I think.

And I do have to wonder if that talent would remain as good without the strong FOSS culture that incentivized me to be there. I worry when I see people like Ian getting laid off too, because there's simply no way that can be about talent. If he doesn't still fit there, that implies a level of change with which I'd have been deeply uncomfortable.

> Google had designed their browser ecosystem and usability around that update model

I don’t know why more Windows software doesn’t use Google’s approach. A low priority scheduled updater is more user friendly than update on launch, and it doesn’t kill boot times like an update on reboot.

From the post: “...I myself was part of 25% of the workforce laid off in August 2020”