It’s no secret that I am very worried about the future of Firefox, and the future of Firefox on Linux in particular. I’m not going to rehash these worries here, but suffice to say that with Mozilla increasingly focusing on advertising, Firefox’ negligible market share, and the increasing likeliness that the Google Search deal, which accounts for 85% of Mozilla’s revenue, will come to an end, I have little faith in Firefox for Linux remaining a priority for Mozilla. On top of that, as more and more advertising nonsense, in collaboration with Facebook, makes its way into Firefox, we may soon arrive at a point where Firefox can’t be shipped by Linux distributions at all anymore, due to licensing and/or idealogical reasons.
I’ve been warning the Linux community, and distributions in particular, for years now that they’re going to need an alternative default browser once the inevitable day Firefox truly shits the bed is upon us. Since I’m in the middle of removing the last few remaining bits of big tech from my life, I figured I might as well put my money where my mouth is and go on a small side quest to change my browser, too. Since I use Fedora KDE on all my machines and prefer to have as many native applications as possible, I made the switch to KDE’s own browser: Falkon.
What is Falkon?
Falkon started out as an independent project called QupZilla, but in 2017 it joined the KDE project and was renamed to Falkon. It uses QtWebEngine as its engine, which is Qt’s version of the Chromium engine, but without all the services that talk to Google, which are stripped out. This effectively makes it similar to using de-Googled Chromium. The downside is that QtWebEngine does lag behind the current Chromium version; QtWebEngine 6.8.0, the current version, is Chromium 122, while Chromium 133 is current at the time of writing.
The fact that Falkon uses a variant of the Chromium engine means websites just work, and there’s really nothing to worry about when it comes to compatibility. Another advantage of using QtWebEngine is that the engine is updated independently from the browser, so even if it seems Falkon isn’t getting much development, the engine it uses is updated regularly as part of your distribution’s and KDE’s Qt upgrades.
The downside, of course, is that you’re using a variant of Chromium, but at least it’s de-Googled and entirely invisible to the user. It’s definitely not great, and it contributes to the Chromium monoculture, but I can also understand that a project like Qt isn’t going to develop its own browser engine, and in turn, it makes perfect sense for KDE, as a flagship Qt product, to use it as well. It’s the practical choice, and I don’t blame either of them for opting for what works, and what works now – the reality is that no matter what browser you’re choose, you’re either using a browser made by Google, or one kept afloat by Google. Pick your poison.
It’s not realistic for Qt or KDE to develop their own browser engine from scratch, so opting for the most popular and very well funded browser engine and strip out all of its nasty Google bits makes the most sense. Yes, we’d all like to have more capable browser engines and thus more competition, but we have to be realistic and understand that’s not going to happen while developing a browser engine is as complex as developing an entire operating system.
Falkon’s issues and strengths
While rendering websites, compatibility, and even performance is excellent – as a normal user I don’t notice any difference between running Chrome, Firefox, or Falkon on my machines – the user interface and feature set is where Falkon stumbles a bit. There’s a few things users have come to expect from their browser that Falkon simply doesn’t offer yet, and those things needs to be addressed if the KDE project wants Falkon to be a viable alternative to Firefox and Chrome, instead of just a languishing side project nobody uses.
The biggest thing you’ll miss is without a doubt support for modern extensions. Falkon does have support for the deprecated PPAPI plugin interface and its own extensions system, but there’s no support for the modern extensions API Firefox, Chrome, and other browsers use. What this means for you as a user is that there are effectively no extensions available for Falkon, and that’s a huge thing to suddenly have to do without. Luckily, Falkon does have adblock built-in, including support for custom block lists, so the most important extension is there, but that’s it.
There’s a very old bug report/feature request about adding support for Firefox/Chrome extensions, which in turn points to a similar feature request for QtWebEngine to adopt support for such extensions. The gist is that for Falkon to get support for modern Firefox and Chrome extensions, it will have to go through QtWebEngine and thus the Qt project. While I personally can just about get by with using the BitWarden application (instead of the extension) and the built-in adblock, I think this is an absolute most for most people to adopt Falkon in any serious numbers. Most people who would consider switching to a different browser than Chrome or Firefox are going to need extensions support.
The second major thing you’ll miss is any lack of synchronisation support. You won’t be synchronising your bookmarks across different machines, let alone open tabs. Of course, this extends to mobile, where Falkon has no presence, so don’t expect to send your open tabs from your phone to your desktop as you get home. While I don’t think this is as big of an issue as the lack of modern extensions, it’s something I use a lot when working on OSNews – I find stories to link to while browsing on my phone, and then open them on my desktop to post them through the tab sharing feature of Firefox, and I really miss it. There’s a longstanding feature request for this one, too, and I suggested something like this could possibly be made to work using KDE Connect.
From here on out, the shortcomings and issues I’ve personally noted fall in the category of minor issues and bugs, mostly stemming from what seems to be a lack of attention in the development of Falkon’s UI. For instance, it’s one of the very few official KDE applications left that hasn’t received the Plasma 6 overhaul to reduce the number of frames, tidy up the spacing a bit, and generally give it a bit of a spruce. This makes Falkon look a bit outdated compared to the rest of your KDE desktop (but not massively so), while also containing some odd UI elements that are quite non-standard. I filed a bug report/feature request about this.
Even smaller bugs are random things like the cursor often not changing to a hand cursor when hovering over a link, favicons in the bookmarks toolbar not updating consistently, or (more annoyingly) screens and laptops going to sleep modes while video is playing. I’m no developer so don’t put too much faith in what I’m about to say, but it feels like things like these are not the most complicated bugs to handle, and looking at the lack of activity in the bug tracker for Falkon, it feels like even within KDE itself, Falkon isn’t exactly used a lot. This is a shame, because other than the issues I’ve mentioned, Falkon is remarkably full-featured and capable.
Think of any feature a modern browser has, and Falkon probably has it. There’s a password manager built-in that can use both encrypted and non-encrypted databases as well as Kwallet, a Greasemonkey-compatible extension, web inspector tools, extensive privacy controls, spell check, address bar search with support for every search engine under the sun, advanced tab management, and much, much more.
And of course, it has a Qt/KDE graphical user interface that integrates better with KDE, both visually and behaviourally, than any other browser – including support for things like the global menubar widget, something Firefox does not support (but Chrome weirdly does). This means Falkon will respect your font choices, colour settings, light and dark mode options, and everything else. Where Chrome and Firefox feel like they’re not at all at home in KDE (or GNOME, for that matter), Falkon is just another KDE application, and looks and feels almost entirely integrated.
Combine all of this with the excellent rendering and performance thanks to QtWebEngine, and Falkon is simply a far more capable and solid browser than most people seem to know. I’ve switched all my machines to it a little over a week ago or so now, and for me, the benefits of having a truly native, KDE-first browser far outweigh some of the shortcommings I’ve had to deal with. Relying on the BitWarden application instead of the extension is a bit more cumbersome, but not massively so. No longer having tab sharing sucks, but since switching to GrapheneOS and its Vanadium browser, I had to give that up anyway.
KDE is on track
My goal with writing this article, talking about Falkon on Mastodon, and filing bug reports – I’ve got a few more extremely minor niggles coming – is to let people know that Falkon is in a far better, more usable, and capable state than most people know. Hopefully, this article will get a few more people to try it out, file bug reports, or even become a contributor to this very capable but seemingly a bit of a neglected official KDE application.
At this rate, I give Firefox (the Linux version in particular) a few more years, at most, before its various advertising and other anti-user additions make it incompatible with the idealogical and license requirements of most Linux distributions, and KDE should be ready for when this day comes. I don’t have any illusions about my influence here – OSNews is relatively small and the inertia to switch browsers is very real and understandable – but I think that with how good Falkon already is, KDE is much closer to not having to rely on Firefox or Chrome than even KDE itself seems to know.
And with how things are going, that’s very good news.
I think you miss the biggest “feature” of Falkon.
It has a Haiku port.
Falkon is one of the current “killer apps” for Haiku and makes the possibility of daily driving Haiku a much more enticing prospect for many. Whilst QT ports are not conduucive to encouraging native Haiku software, the reality is that Haiku needs apps, and if Falkon provides a stable and compatible web browser for it, then it’s a damn sight better than nothing.
There is also a (native i believe) modern Firefox port in the works as well. However, the particular developer of the port has a habit of making things “work” and then leaving the “maintain” bit to others.
I came here to say this. Falkon, more than any other app, makes it possible to use Haiku in the real world. It is vastly better than WebPositive and the other Haiku alternatives.
I’d also like to add that Falkon is standard on Genode for largely the same reasons: that operating system also facilitates Qt ports. And with the other Qt software ported so far it also makes Genode close to “daily driver” albeit not so much as Haiku.
IMHO it’s very short sighted to worry about the future of Firefox in a way to predict its full demise is coming soon and then to provide a solution in terms of Chromium based web browser. When it comes to Chromium based web browsers the future is already written and it’s called Chrome. The idea Falkon can in any meaningful way take on Chrome is a pipe dream.
Exactly this. I’d sooner switch to a Firefox fork (Librewolf comes to mind) than a Chromium-in-disguise.
The point I make in using Firefox is supporting a different engine, diversity in the web platform. Giving up is not an option (yet.)
Furthermore I expect Firefox Sync to be able to function.
If you are looking to check-out Firefox derived browsers, give Zen a look.
> This effectively makes it similar to using de-Googled Chromium.
So why not just use Ungoogled Chromium instead of Firefox? Especially since unlike Falkon it offers binaries for Windows as well.
Falkon has binaries for windows and had so for many years. It also has binaries for Linux, BSD, Haiku and many more.
yep
2019-03-22 13:49
“It’s not realistic for Qt or KDE to develop their own browser engine from scratch”
Not only is it realistic, they’ve done it before, and almost everything is based off that work.
Khtml was a KDE built engine that was forked into WebKit (Apple) and later Blink (Google)
The project then died because everyone moved on to those projects instead. (Also Apple/Google directly hired key members of the team and paid them, which definitely accelerated the demise of the project).
I think you just proved that it *isn’t* realistic. As the project go so far and before it was really ready was raided by Apple who co-opted it and created something really quite different that the community no longer had control over. And of course complexity only increases over time, creating a browser rendering engine that matches firefox, webkit or chrome now is a lot easier than matching firefox or ie back in 2002.
I guess it’s more “will you work for money or for free”.
Khtml wasn’t That great when Apple chose to use it. But it wasn’t gecko, trident or opera.
Money, time and money gave it focus.
We can see between ladybird and servo how quickly a new engine can be built, for (in the grand scheme) very little money
Again, your own comments seem to be proving you wrong. LOL Servo has been in development for over a decade now and no real browser exists that uses it. Ladybird will take four years to get to alpha in 2026 but parts were worked on for longer than that. How is that quick?
@MJ
5 years or so to create a “real” web browser seems pretty fast to me. Nobody is disagreeing that they are massive undertakings. Ladybird is certainly proving that it can be done though. We will see what “Alpha” means when we get there. Ladybird is hardly ready for primetime but you can, as an example, surf OSnews with it quite nicely already. For a few weeks, it was a ritual of mine to compile the latest version of Ladybird everyday and then to come here. I was able to read most articles as well and login and comment just fine.
I hope Servo or laydbird mature to become viable alternatives. That’s what I was hoping Thom was hinting at.
QtWebEngine is notorious for always having a lot of unfixed security issues that have been fixed in upstream Chrome. Qt release cycles just aren’t as fast as Chrome’s. It’s not meant to be used as a browser, but to render trusted content, think of for example of the Steam app rendering its own store. Using it for a browser to render untrusted content, you’re basically giving every website the ability to execute whatever code they want on your computer. I really hope nobody gets owned because they followed your advise from your article to use it as a main browser.
Then you’re also switching to Chrome the rendering engine, further increasing the market share of it, helping it to gain complete dominance and hence giving full control of the Web to Google. Firefox market share is the only option to force websites to support more than just Chrome’s engine. Safari might be another.
Firefox might have an uncertain path ahead, but now is not the time to abandon ship. As doing that now is just handing the full control over the entire Web to Google. Wait until Firefox *actually* does something that makes it impossible to keep using it. At that point, there will certainly be forks.
And if you absolutely don’t want to use Firefox, at least use something with WebKit and not Chrome.
Falkon switched away from WebKit not that long ago.
WebKit is a credible Chromium competitor currently as it is used by Safari, the number two browser by market share. GNOME Web ( Epiphany ) still uses WebKit. However almost nothing else uses does. WebPositive ( Haiku ) uses it but it is not very good.
The Qt toolkit moved from WebKit ( QtWebKit ) to Chromium ( QtWebEngine ) when Qt5 released. That is what caused Falkon to switch.
So, the solution for the downfall of Firefox is switching to yet another Chromium browser which is unavailable on Windows, macOS, and any mobile OS.
Yeah, its pretty sad. It is the way it is, and I don’t fault Thom for choosing it. But I was really hoping that he had magically found some other non chrome/webkit solution.
Falkon has been available for windows for many years. You can get the 3.2.* releases from the attic.
https://download.kde.org/Attic/falkon/3.1/
I was about to reply to your other comment here for a link to this mysterious Windows version.
Thanks for confirming it is not actually available for Windows (anymore). No cares about a release from 2019. WTF? LOL
You can still build from source just fine. I just tried it.
is this a joke? Why would one build a mediocre at best browser from source on Windows and for every new update? I hope you meant at least it’s possible to compile the current release source on Windows and not what you linked here.
We collectively painted ourselves into a corner. When Google came with Chrome and released most of it as FLOSS Chromium, we believed they “were the good guys”. Mozilla was on top of the world as the alternative browser. Using Chrome/Chromium couldn’t harm, right? It was just an addition to the browser pantheon.
Flash forward and Google is the current big, bad company and they masterfully maneuvered Chrome to be their moat around the search/advertising business. Mozilla is on life support offered by, drumroll, Google and most independent browsers are now nothing more than a skin on top of Chromium.
Maybe Servo can be the knight in shining armor and save us from the Blinkpocalypse. Let’s hope so. I just hope we will be smart enough not to fall for the same trap again. If at any time a big corporation co-opts and forks Servo, maybe let that product wither at the vine. We can’t trust commercial interests to serve our needs, unless the situation is like the Linux kernel; too many participants for anyone to hijack the process.
Well Linux is a dictatorship. So ultimately one man decides.
The comment you replied to was talking about the state browsers in general not just Linux. Regardless, you are misinformed as Linus has nothing do with browsers on Linux.
I am not sure who these people are you are talking about that thought Google “were the good guys” but OK. LOL
The problems with Firefox has nothing to do with it’s rendering engine so I am not sure how another rendering engine by itself is going to save us at all. This may get my comment deleted but you can look into Ladybird which is a full browser built from scratch not based on Chromium.
At the end of the day the 3 top browsers are all based on Chromium and those browsers come from 3 big for-profit tech companies that have their own platforms so good luck in getting traction with an alternative. The regulators even look to be useless as they are talking about making Google sell Chrome while a better solution would be getting Chromium out of Google’s control.
During the oughts Google was not yet considered evil incarnate. The browser landscape was completely different. There were multiple browser engines. KHTML, WebKit, Trident, Gecko, Presto. It made standards compliancy practically mandatory. It also wasn’t a mono culture, where an addition to the rendering engine automatically makes it a standard. In that way, all the different rendering engines made W3C have importance.
Now that the only independent hold out (with impact) is Apple’s WebKit, Google can just practically do whatever they want with Blink and have it be what the web runs on. That is what the problem with de facto one rendering engine is. Google calls the shots.
The trouble with Firefox is indeed not its rendering engine. The trouble is that Mozilla is largely irrelevant in today’s world and it limps along on Google’s money. Firefox is the diversion that Google uses to give the illusion that there is competition in the browser landscape. For non-Apple users it seems all roads lead to Google’s hellscape.
Another rendering engine with a sufficiently large installed base can take away some of the power Google now has with its Blink monopoly (outside of the Apple ecosystem). It would compel Google to at least keep compatibility with the other engine. Two incompatible rendering engines, both with enough installed base, would split the web and that isn’t in Google’s interests.
No idea why you are explaining browser history to me. I was there the whole time. LOL I never switched form Firefox to Chrome. Did you? Firefox still works just fine and a rendering engine from another trillion dollar tech company that is mostly just available on a proprietary, closed platform is not of much interest to me and would just be a setup to repeat history. No one actually thinks there is “competition in the browser landscape” either. We should be trying to push Mozilla to focus on and turn around Firefox as that’s the only browser with any significant market share outside the big three but good luck.
A company can do a lot of good and contribute to open standards and opensource projects but no one should be surprised at the end of the day it’s all about their bottom line for them and expanding it.
I’ve tried Falkon quite a bit over the years and I really like it. The one thing that keeps me from using it as my everyday browser is that it doesn’t have extensions with the capability to do per-domain script blocking. Without having uBlock Origin at a minimum, or preferably uMatrix/eMatrix, or worst case scenario at least NoScript, it’s not something I’m willing to use daily. I’ve stuck with Pale Moon because of its continued access to uBlock and eMatrix, and I really don’t see much web rendering difference between Falkon and Pale Moon.
Since Pale Moon has been independently developed with its own Goanna browser engine for quite a lot of years now, the downfall and destruction of Mozilla should have little to no impact. However, Falkon will still be reliant on the good will of Google’s chromium development, and as we’ve seen with the enforced imposition of Manifest V3, that good will is not assured.
andyprough,
It’s good to take stock of our options, but it’s also important to consider just how independent some of these alternatives actually are. Google controls a lot of the code, but even when they don’t they still have tons of leverage over the web. Like blocking access on youtube. I don’t know if google will try it again, but they could push for web DRM again. Services like netflix, banks, governments, etc could end up requiring said DRM as a condition of service. Whether we like it or not this would affect all of our favorite alternative browsers. I don’t think this is a stretch, many companies already do this with their mobile apps and they would support this model for the web too. The bigger question is whether the tech giants can eventually find a way to push this through and normalize it. At least on this front mozilla had more resources to lobby for consumer interests. But if they should they fail, then the smaller alternatives in the long tail would likely end up carrying even less weight on open standards than mozilla.
These are good reasons to use Falkon, I think what you are saying supports Thom’s position.
With the way the entire Alphabet company is on the chopping block in US Federal court right now, I’m thinking Google is not currently in a position to make even more monopolistic moves like building locked-down DRM into all kinds of website access. But it could happen like you say. I think Thom is right, we should all be looking at something like Falkon as at least our secondary browser. I would like Falkon better if it had extensions, but Thom is right that it’s certainly a capable browser that aesthetically fits in perfectly into a GNU/Linux environment.
andyprough,
I’d agree that’s the case right now, But timing is everything. Come next US administration, federal regulators themselves may find themselves on the chopping block. Trump is a notoriously corrupt politician and so if big businesses stroke his ego, these restrictions will probably go away.
Net neutrality is a prime example of how malleable rules are depending on the administration.
https://www.reuters.com/technology/us-agency-vote-restore-net-neutrality-rules-2024-04-25/
I’m not ready to abandon FF yet, but it seems inevitable that mozilla are going to experience yet more layoffs and shrinking marketshare. I have little faith in the community helping mozilla stay afloat given the harsh words being written about them…from people who should be their base supporters no less.
>”I have little faith in the community helping mozilla stay afloat given the harsh words being written about them…from people who should be their base supporters no less.”
That’s always a good way to get people to support a software project – try to shame them into it. Hope that works out for you.
andyprough,
I do not understand…what’s up with this response? There really was no expectation on my part that things will work out if we shame people. Quite the opposite the way things are going. If you disagree with me do you mind clarifying?
>”If you disagree with me do you mind clarifying?”
I think I mis-read what you said. You said, “little faith in the community helping mozilla stay afloat”, I read that as “little faith in the community helping Firefox stay afloat”. I agree with you, there’s probably almost no interest in the community for keeping Mozilla afloat. However, I think there will be a lot of developers volunteering to keep Firefox afloat once Mozilla is out of the way. I think Firefox has a bright future as a community project once it is un-tethered from Mozilla and from Google’s overbearing influence.
And so we’ve come full circle. 🙂
KDE’s browser Konqueror used the KHTML engine.
Apple took the KHTML engine and re-worked it into WebKit for Safari.
Google took WebKit and re-worked it into Blink for Chrome.
Google released Blink as part of Chromium.
Chromium’s version of Blink was imported into QtWebEngine.
And KDE’s new browser Falkon uses QtWebEngine.
Microsoft is talking about replacing the browser with AI interfaces. This may all become irrelevant pretty quickly.
Already we past the browser-peak – more and more people interact with internet via native apps on smartphones.
I wish Falkon was Gecko-based. Don’t care about chromium wrappers.
Thom,
Based on your criteria, why not Konqueror or Angelfish on KDE?
Twenty-ish years ago, when KDE was my primary DE, I relied on Konqueror quite regularly as it obviuoulsy integrated with KDE well, had great FTP capabilities, and it was easy to work with downloaded files since it was also my default file browser back then. I did eventually move away to other DEs and WMs and therefore moved away from Konqueror, but it still seems pretty solid when I installed it recently and played around with it.
Every couple of months I try out Plasma again on my phone, but always end up back on Phosh. The default browser for Plasma mobile on several operating systems is Angelfish, which is always a highlight of the Plasma experience. It seems quite capable as well as I was able to stream from my friends Plex TV server and run my web based CAD software no problem with it. In fact, I’m posting from Angelfish on my laptop right now.
All that said I am not ready to switch from a Firefox quite yet. It integrates nicely with my email client of choice, Thunderbird. Maybe I would feel different if I were a KDE user, but currently I enjoy the Linux P0RN of Hyprland. I do use a number of KDE/Plasma tools though. As an example. my default file browser is Yazi and I have it set to launch in Konsole despite my default terminal being Alacritty.
Just curious as to why Falkon over those other two options.
I always adored falcon browser, thank you for this post.
Free AI Tools
I’ve found Falkon to be a good balance of features and light weight for low-end hardware and virtual machines. It’s more capable (and compatible!) than NetSurf or Dillo, and lighter than Firefox or Angelfish. It also runs well under LXQT, which I like to use on that low-spec (virtual) hardware for the same reason.
In my case the two extensions I miss most are the KeePassXC integration and Floccus, which syncs bookmarks across multiple Gecko & Chromium browsers. I can sort of get around the first with auto-type, at least
While I do acknowledge that the web landscape is radically different today than it was 20 years ago, I chuckle a bit at the statement that “KDE cannot develop a browser from scratch”, considering that WebKit, developed by Apple, started as a fork of KHTML (rendering engine of Konqueror and KDE in general), and is probably the sole reason why WebKit ended up being open source. I’d say KDE is responsible for both Safari and Chrome rendering engines. I agree it’s a bit far fetched statement, but history needs to be remembered regardless of my personal hyperbolas :D.