A lot of Firefox's loss of traction seems to be related to 1) development tools not reaching people's expectations, 2) Gecko cannot keep up with WebKit's speed of development, 3) general perceived performance and ui issues in Firefox.

Mozilla doesn't have unlimited resources, and from an external observer, it seems that new projects, such as B2G/FFOS, and other products/services would benefit from a little bit more attention.

What seems to happen right now, is that a great deal of resources inside Mozilla are dedicated to create duplicate, parallel work to what's being done out there. I'm referring to 1) and 2), the platform on which Firefox and many other products depend on.

A Gecko brain transplant would be hard though.

Arguments against:

1) Webkit is fragmented
2) Switching would be so expensive it'd eclipse any benefits
3) Webkit is Google/Apple, therefore proprietary
4) Shared components stagnate, we need different implementations
5) Standards happen when there are multiple implementations

Arguments in favour:

6) Webkit is solid, fast-moving
7) Switching would ultimately reduce effort duplication, release resources
8) Webkit is open-source, forks are possible, Mozilla would still be able to implement its own version of spec X, test it in the field, push back to Webkit and have it adopted by other vendors effortlessly instead of requiring a new implementation.
9) The value is not in the rendering/etc engine, the value is in the experience, therefore UI and services should have more focus.
10) Gecko is rowing against the stream on mobile, will have to go through a lot to be where Webkit is today, a lot would be gained if focus could be shifted away from the barebones onto marketing, ui, ux, docs, dev relations for FFOS.

No. Diversity is good.

I agree with Rui. Diversity is good, but it's better when it competes at the same level, such as when Gecko/FF were setting the new trends and ahead of everyone else.

I think Luis basically nailed it on his first post when he said Mozilla doesn't have unlimited resources.

I'm just speculating here, but I assume Mozilla could benefit from some better reorganisation and maybe some focus on some specific projects to keep them more competitive with other implementations.

I see where Mozilla wants to go with B2G and FF mobile versions, and I think it's necessary because everyone knows that's the "next big thing", but I can't help but feel that some of their teams are being stretched too thin and can't keep up with the pace of other projects.

To clarify, diversity wouldn't be threatened.

I believe the web could move faster if the implementation costs were reduced by using a shared codebase.

WebKit has moved to a modular codebase**, and now features WebCore modules. (probably an ongoing effort.)

** trac.webkit.org

Features such as LocalStorage are contained in their own modules, making it easier to come up with alternative implementations, test them in the wild, and be adopted by other browser vendors.

Also, is hardware integration part of the WebKit project or something owned by each vendor (google)?

No. A monoculture (even if open-source and modularized) threatens/stifles innovation. Please don't.

The only question I have to contribute is that I'm not sure what you meant that WebKit is proprietary to Google and Apple.

At the same time, I could see the benefit to Mozilla being able to focus on differentiating Firefox instead of keeping up with Safari and Chrome, but I don't think that we want 75% of the popular browsers out there to use the same codebase.

@Rob: it's not my personal opinion, WebKit is seen by many as owned by one or both of those companies, although we know its an open source project anyone can contribute to.

I still don't see a problem with sharing the code base and continue to see it as an advantage.

Opera seems to think the same and has recently switched to WebKit/V8, becoming a contributor and focusing on innovation. (I've added @brucel to this branch in case he wants to add something)

Today's news have made me think a bit more about this, and while I haven't changed my opinion regarding not liking monocultures, I think this Opera switch isn't bad and people are making a big deal out of it.

Opera have a fairly low market share (ok, I know about mobile and such) and I don't know the size of their team, but I believe they can help a lot more by putting their efforts into Webkit than competing with it.

However, I think we still need IE and Firefox to be a counterpoint and competition to the Webkit world.

Imagine one of the contributing companies to Webkit comes up with a new feature and they ship it in their Webkit version, but it doesn't get merged into the Webkit upstream because other contributing companies don't want it.

If competing engines decide this feature should exist and also implement it and push for it, suddenly there's more overall support for it and it might bring other Webkit contributors to accept it's existence and standard bodies to agree it should be standardised and part of the spec.

In a monoculture, even with various contributors, the politics of these situations will always be very different it could lead to even more Webkit fragmentation, which personally I think is a worse problem than competing rendering engines.

I see your point.

1) is webkit a culture?

I actually read your paragraph as 'If different contributors' instead of 'competing engines'. Wouldn't the benefits still be present?

I can see WebKit fragmentation problem becoming worse, especially because of how standards evolve and we use them as they're implemented.

Currently, and as far as I know, this kind of fragmentation currently revolves around features, performance, updates, not behaviour. ( interestingly enough, webkit is the only prefix that lacks association with an actual vendor, and old webkit versions do cause behaviour changes )

And also: up until recently we had different implementations of WebRTC, preventing interoperability between FF&Chrome. Fragmentation is already here.

2) do you think we can establish a parallel with the linux kernel, or better: linux distributions?

In case of the kernel, there are many contributors, forks - most popular being the one used in Android and that got some criticism for not pushing back to upstream.

Just a couple of updates:

@brucel and @jeresig both chimed in on the subject and appear to be of the opinion that this isn't the same as IE and poses no threat to innovation.

The Articles:

"Bruce Lawson on Opera’s Move to WebKit" - alistapart.com

"WebKit is the jQuery of Browser Engines" - ejohn.org

One more from @jaffathecake

plus.google.com