4 ms·

> Losing the browser would be a catastrophe, losing the rest would not even register Here's a sort of thought experiment, which I really don't know the answer

> Losing the browser would be a catastrophe, losing the rest would not even register Here's a sort of thought experiment, which I really don't know the answer to. How much of Firefox development (by hours, lines of code, commits, pick your metric) is done by Mozilla employees as part of their day jobs? And how much of that development is ~useless for the maintenance and improvement of the core browser? (Stuff like Pocket and the "discovery" feature on the new tab page, all the time spent on the URL bar changes, etc etc.) It might turn out to be the case that most, or even all of core Firefox development could be done by the community. At the very least I'm willing to put out there that maintaining a web browser does not require half a billion dollars a year. (It's rather hard to imagine that 100 full time devs making $250k a year each couldn't handle basically all of the core development, for example.) The problem with Mozilla is that it's a business, for all intents and purposes. With an open source project, as long as ongoing development is financially maintainable you don't have to care about marketshare. Mozilla cares about marketshare because it affects how much Google will pay them for search traffic (and how much they get from in-chrome advertizing). When Oracle proved derelict in its duties managing the development of OpenOffice, the Document Foundation was formed, largely by people who were already spending their time on OpenOffice. They forked OpenOffice into LibreOffice almost overnight. Major longstanding pull requests were quickly merged. New people who previously hadn't wanted to contribute stepped up to support a community project. And the result has been all to the good! The Document Foundation still manages the project, and in 2019 did so with a total income of less than a million Euros. To me this is proof that large open source software projects can quietly evolve into two camps: one is the employees of a governing corporation, and the other is the community. And the community might be doing most / all of the core work required to keep the project running, and the employees are mostly doing stuff that benefits the company. And suddenly there's a community fork - and it just works, without any hitches. I don't know if this would be successful in the case of Firefox, but I expect (and hope) that people would try if there was ever a hint of Mozilla abandoning it. (For another example, consider KDE e.V., which manages an enormous open source project on a budget of under 600k Euros a year as of 2018.)

InfiniteRand 6y ago
There are a lot of dead or mostly dead browser projects out there which makes me think community alone is probably not enough (as a comparison, for a C/C++ programming IDE there are enough moderating active projects out there that I'd say community is enough to make an IDE). On the other hand, 10-20 years ago, there were a lot of alternative browser projects that were moderately active and able to keep up with 90% of the web.
jpetso 6y ago
I was part of the BlackBerry browser team in the days of BlackBerry losing the mobile market. We maxed out somewhere between 50 and 100 people working on the mobile browser, and that doesn't include business/marketing because it was just a part of the overall smartphone product. Maintaining a browser is an immense undertaking. With 50 full-time contributors, you may just about be able to provide a functional chrome to Blink (then WebKit) that doesn't suck too badly and keeps up with the performance and web standards requirements that Google in particular keeps pushing forward. 100 people is in no way enough to make both production-quality browser chrome and browser engine on your own, become the cutting edge for performance and privacy improvements, keep up with and participate in standards development, plus whatever auxiliary supporting efforts are necessary when your browser is the main product as opposed to supporting a cashflow-rich ad or app store business. Sure, you can build a Chromium derivative for early adopters or crypto enthusiasts with a handful of people and a bunch of VC money. But what Mozilla does is more important, and has a much larger scope. They may survive with 100 people for a while, they would be a far cry from being relevant though.
- jacquesm 6y ago
  That's actually useful information. There are 750FTE at Mozilla. How many of those do you think would it take to do the best job possible on the browser. That would give one way of calculating a budget.
  - jpetso 6y ago
    Hard to say, and I suck really bad at estimates. If I low-ball it at maybe 200-300, that's likely off by a factor or three. A possibly old but close enough page on chromium.org lists the following teams: Animations Team Binding Team Device Team DevTools DOM Team Ecosystem infra Input Team Layout Team Paint Team Rendering Core Speed Metrics Team Style Team Web Capabilities (Project Fugu) Worker Team and that looks like it's only focused on Blink, not including the actual chrome (desktop/mobile apps) or V8 which would both be substantial efforts as well. Who knows how much technology is pulled from other teams at Google, like video encoding for WebRTC, server performance collaboration for the SPDY/QUIC effort, low-level storage for IndexedDB, devrel with buggy websites or certificate authorities, security audits and whatnot? For doing the "best job possible", you can probably blow up team size by a lot, there are numerous tasks that would benefit from more attention and I'm sure Mozilla are already making hard trade-offs at their current staffing levels (which obviously include a lot of people outside of Firefox engineering).

InfiniteRand 6y ago

There are a lot of dead or mostly dead browser projects out there which makes me think community alone is probably not enough (as a comparison, for a C/C++ programming IDE there are enough moderating active projects out there that I'd say community is enough to make an IDE). On the other hand, 10-20 years ago, there were a lot of alternative browser projects that were moderately active and able to keep up with 90% of the web.

jpetso 6y ago

I was part of the BlackBerry browser team in the days of BlackBerry losing the mobile market. We maxed out somewhere between 50 and 100 people working on the mobile browser, and that doesn't include business/marketing because it was just a part of the overall smartphone product. Maintaining a browser is an immense undertaking. With 50 full-time contributors, you may just about be able to provide a functional chrome to Blink (then WebKit) that doesn't suck too badly and keeps up with the performance and web standards requirements that Google in particular keeps pushing forward. 100 people is in no way enough to make both production-quality browser chrome and browser engine on your own, become the cutting edge for performance and privacy improvements, keep up with and participate in standards development, plus whatever auxiliary supporting efforts are necessary when your browser is the main product as opposed to supporting a cashflow-rich ad or app store business. Sure, you can build a Chromium derivative for early adopters or crypto enthusiasts with a handful of people and a bunch of VC money. But what Mozilla does is more important, and has a much larger scope. They may survive with 100 people for a while, they would be a far cry from being relevant though.

jacquesm 6y ago
That's actually useful information. There are 750FTE at Mozilla. How many of those do you think would it take to do the best job possible on the browser. That would give one way of calculating a budget.
- jpetso 6y ago
  Hard to say, and I suck really bad at estimates. If I low-ball it at maybe 200-300, that's likely off by a factor or three. A possibly old but close enough page on chromium.org lists the following teams: Animations Team Binding Team Device Team DevTools DOM Team Ecosystem infra Input Team Layout Team Paint Team Rendering Core Speed Metrics Team Style Team Web Capabilities (Project Fugu) Worker Team and that looks like it's only focused on Blink, not including the actual chrome (desktop/mobile apps) or V8 which would both be substantial efforts as well. Who knows how much technology is pulled from other teams at Google, like video encoding for WebRTC, server performance collaboration for the SPDY/QUIC effort, low-level storage for IndexedDB, devrel with buggy websites or certificate authorities, security audits and whatnot? For doing the "best job possible", you can probably blow up team size by a lot, there are numerous tasks that would benefit from more attention and I'm sure Mozilla are already making hard trade-offs at their current staffing levels (which obviously include a lot of people outside of Firefox engineering).