Getting a large enough group of contributors is hard. For Linux this works since it was built over decades and contributors come from variety of backgrounds (students who want to learn, hardware companies who want to have support of their hardware, users who want it to be fast, companies which want to make it sort of a product (distributors etc.))
For "end user" software like firefox this is harder. In the kernel a contributor can start by providing a self contained device driver. Such a concept doesn't exist in a browser in similar form. From such a driver the kernel contributor can grow into neighboring subsystems.
Google decided to go with Chrome. Microsoft jump on their ship and is unlikely to switch, again.
Who else could push this? Facebook (would we like that!?), Twitter, Amazon?
I'm most probably somewhat naive here, but if you could fork the latest release, remove all Pocket, VPN, whatnot stuff and just give me that as a build, I would be totally fine with it.
I love the core product, I really don't need any of the fluff around it.
>, but if you could fork the latest release, remove all Pocket, VPN, whatnot stuff and just give me that as a build, I would be totally fine with it.
You've overlooked the main idea of the parent post. Let me try to restate it another way...
You can't just fork a web browser's source code and be done with it. A web browser needs constant ongoing new programming coding to keep up with the ongoing changes in the web ecosystem.
Already, we've seen the complexity of new web standards make Opera give up on their Presto web rendering engine and Microsoft abandon their Trident engine for Internet Explorer. Both companies switched to Chromium source as a base to save money and resources.
So no, you eventually would "not be totally fine with it" -- because your forked browser would eventually be useless without a big team of programmers to maintain it.
As examples of web browsers quickly becoming obsolete, I tried Opera 12.18 (last old 2016 version with Presto engine) and here are many problems I encountered:
- Google Maps -- Zooming in and out makes everything blurry. Street View hangs the browser. Opera is missing WebGL acceleration that today's browsers have
- godbolt.org -- compiler explorer online C++ website is broken with a blank screen and doesn't show any code panels
- chase.com -- that bank doesn't allow sign in with old Opera browser
- reuters.com -- images on news stories are blurry and don't load correctly
- various websites with newer TLS encryption protocols break because they don't exist in Opera
Nobody wants to take a fork of the Presto engine and expend 1000 man-hours to fix all those problems. Same would happen with a hypothetical fork of Mozilla Firefox. You still need an active programming community to keep up with evolving web technologies. Again, if Opera and Microsoft (with its billions) gave up, it should give an idea of how daunting it is.
You're right. I was thinking that "add-ons" like Pocket might be modularized enough so that removing those parts would be easy with every release, maybe even (semi-)automated. Like a "Firefox Light" version.
Maybe this could even be maintained by Mozilla, I would not care in that case. Just the bare core browser. You can always choose to install "the normal" Firefox if you want the "full experience", but for me a browser is just another tool, I don't need pocket and alike.
Yeah, maybe it's a little piece and nothing to worry about. But what will they come up with next? Maybe at one point they`ll integrate the "Mozilla VPN" functionality directly into Firefox. Or any other fancy thing they thought the world would need... So I guess it's more about software simplicity, but I find that helpful. I think it also helps developers if they have a clear goal ("build a great browser!") and don't need to take care about integrating side-projects and stuff.
See the other comment: I would have hoped that removing "modules" like Pocket would be easy enough so it would be feasible to maintain a "Firefox Light" version.
> Getting a large enough group of contributors is hard.
Not really. This is a thought-terminating cliche in the form of a just-so explanation. The group of volunteer contributors who were underpinning Mozilla's success at its peak was quite large and healthy, complete with its own governance model. But then Mozilla Foundation essentially handed the keys over to the Firefox subproject, which in turn made a bunch of questionable hiring decisions, including people that deeply affected the project at an organizational level.
The really perverse thing is that a very similar thing had already happened once before—with Netscape. It was those very circumstances that led to Mozilla's original governance model as an insurance policy against any one corporate group accumulating too much power after getting it into their heads that they were the real reason for the project's upward trajectory and the only thing they needed to do to really get things on track was wrest full control of the golden goose. Mozilla's "failure" began when every random yahoo who managed to get through the company's hiring process was ipso facto part of a privileged class and anyone not on payroll became second-tier consideration at best (or more realistically, seen as a nuisance).
This "failing", of course, is the kind of failure that grew revenue to half a billion per year even as Firefox was getting thrashed, so whether you consider that a failure or success will come down to principles. If your idea of success is tied to the organization's charter, its openness and approachability, enabling people to shape the future of the Web even when not on the payroll under a browser vendor, and its "market" adoption and the tractability of developing an independent browser engine, then Mozilla is a failure. If, on the other hand, you're predisposed to seeing high revenue as a success, then Mozilla has been winning up to the point when it will inevitably fail due to unsustainability and its approach of salting the earth around it, pulling up the metaphorical ladder, etc.
> For Linux this works since it was built over decades and contributors come from variety of backgrounds (students who want to learn, hardware companies who want to have support of their hardware, users who want it to be fast, companies which want to make it sort of a product (distributors etc.)
That's a pretty accurate description of Mozilla for a large part of its history pre-FirefoxOS. Mozilla was incredibly effective even before the number of people getting paychecks from anything called "Mozilla" had reached the low hundreds.
Yes, Mozilla had/has non-employed contributors. But I (wrongly?) never had the impression the amount of contribution was thaaaat notable (while individuals like Ben Buksch, whom I remember from that time, got some attention)
How hard it is, we can see even within Mozilla related products: Since Mozilla stopped the Mozilla Suite (SeaMonkey) hasn't seen much uptake, and Thunderbird is mostly stagnating as well.
Agreed in principle. It's my primary browser and it seems to me that it's already in good shape (technically, and feature-wise). And the project is not starting from scratch.
So if the financial incentives aren't there yet, and the only remaining value that Firefox provides is privacy and user freedom (along with good but not perfect performance), I'd imagine the FSF could be interested in managing something like this. And can maybe start with part-time volunteers.
This makes it even harder from my p-o-v. Most contributors want to create something new and not maintain old legacy, grown over a few decades. Diving deep enough is work and might happen once in a while for a bug which is a pain for somebody curious ... but tough sell.
Getting a large enough group of contributors is hard. For Linux this works since it was built over decades and contributors come from variety of backgrounds (students who want to learn, hardware companies who want to have support of their hardware, users who want it to be fast, companies which want to make it sort of a product (distributors etc.))
For "end user" software like firefox this is harder. In the kernel a contributor can start by providing a self contained device driver. Such a concept doesn't exist in a browser in similar form. From such a driver the kernel contributor can grow into neighboring subsystems.
Google decided to go with Chrome. Microsoft jump on their ship and is unlikely to switch, again.
Who else could push this? Facebook (would we like that!?), Twitter, Amazon?