Desktop Linux has a Firefox problem
osnews.com
osnews.com
Is it really the best solution to create a fork?
Why not invest in people to directly work on Firefox?
The author even mentions a few paragraphs before that sentence that maintaining a browser is hard:
> The problem here is that making a capable browser is actually incredibly hard, as the browser has become a hugely capable platform all of its own. Undertaking the mammoth task of building a browser from scratch is not something a lot of people are interested in [...]
And who guarantees that the fork will further exist without weird funding?
EDIT: sorry, after reading my comment again, I think it sounds a bit snarky. English is not my native language, and I don't know how to phrase these questions better, so they don't sound that way.
To me, Mozilla is like the Wikimedia foundation- they take in a lot of money, but it seems like very little actually goes to the projects people think of when they hear the name.
Maybe that isn't a bad thing- surely there are worthy projects they could be investing in, in addition to FF or Wikipedia - but supporting them isn't as simple as offering more bodies in seats, because lack of money and talent isn't what is holding them back.
Donations all to to Mozilla, and if the article is to be believed, Mozilla won't use the extra money or FF for Linux.
Hell, in the past we've seen Mozilla divert funds away from Firefox to support unrelated stuff.
There is currently no way, right now, to give financial support to Firefox development without forking it.
This is not true. There is no way to fund Firefox development by giving money to Mozilla, but you can send as much money as you want to Firefox contributors, incl. hiring someone to fix bugs full time, just like any other project.
I don't have any money giving Mozilla money to go mismanage.
However I would be willing to schedule recurring donations for Firefox and Thunderbird development.
How the author arrives at the conclusion that the answer is to create another browser is beyond me.
Advocate for Firefox. Invest in Firefox. There you go.
To summarize: Firefox's shrinking marketshare is an existential threat to Mozilla. The vast majority of its revenue is tied to the search engine deal with Google that the author describes as quasi-charity. Linux is clearly a lower priority than Windows and Mac (and the author even says that it's logical for Mozilla to prioritize Windows/Mac). So if there is a substantial drop in revenue (like the Google deal falls apart or Google just decides to pay Mozilla less) then Linux will bear the brunt of the reduction in resources.
The author sees reliance on Mozilla as a risk
yeah, so the best thing to do is to try to reverse that. the article just takes it as given that mozilla/firefox will probably fail.
and if it does fail, there's a lot of linux users out there installing chromium on their desktops that will have helped it along.
its so weird how the browser is arguably more important than the operating system at this point, but open source nerds don't even see the problem in shitting on firefox and using google's browser instead. that attitude could be changed.
> And who guarantees that the fork will further exist without weird funding?
I think creating a fork could be possible, but I don't really see how this solves the core problem that the author is worried about and how this is the "ideal solution".
> Is it really the best solution to create a fork?
This is funny considering that WebKit was originally a fork of KHTML.
Linux is not an operating system for the unwashed masses, it's a specialist operating system for more technical oriented folks. It's less polished in many areas and requires some jumping through hoops at times as we all know. In return, you get a vastly better user experience if you're the right kind of user.
Whenever I am placed in front of a standard Windows install, I'm shaking my head in astonishment and wonder how people put up with that. Yet, if I were a standard Windows user placed in front of a linux machine, I'd probably do the same.
We're just a different species of users. If linux users cared as much about browser specifities as they did about other things, I'm sure they would already be fixed. But the truth is: as a long-time linux user I can state that Firefox is just fine. Really, it is. For me. And probably for many other happy linux users too.
Does that mean if couldn't be improved? Of course not. But show me the software for which you couldn't say the same. Okay, maybe Emacs. Just kidding.
I remember Linus and Luke from Linus tech tips on YouTube making a series about the Linux desktop, in the end they both had a bunch of very constructive criticism, things that the "more technical oriented folk" would just brush off as minimal issues that can be dealt with it at the end of the backlog. But those issues are actually quite important, and if there's a chance that Linux can be open to more people, then let's do what we can to make that chance a reality.
Fans of the browser (starting with the guy in my mirror) need to start subscribing to services like their VPN or something to give Mozilla a better financial footing.
> The best and most visible example of that is hardware video acceleration.
That's not "Desktop Linux has a Firefox problem", that's "Desktop Linux has a problem". hardware acceleration was always a pain to get working in linux, as a subset of the larger problem of linux video drivers having varying levels of support for the plethora of features that video cards need to do these days.
Things like this ultimately led me to my current Linux usage: Windows host, linux vm, fullscreened on its own virtual desktop.
If I need to do something that linux wasn't going to do efficiently, I do it in the windows host. Otherwise, I do it in the linux vm.
Anyway, I wonder if any of the problems this person is pointing out are going to be made better or worse by wayland
Media players like mpv supported VA-API for video acceleration for many years and that just worked for me without any issues on all my computers.
It is clearly a Firefox issue that it did not support API that was supported by many Linux-native software.
Windows 10 + WSL
I thought an AMD + Radeon laptop would be a great improvement. Evidently I have the Reverse Midas Touch. Also garbage. Sadly, FULL hardware/peripheral support requires Windows or Mac, happiness everywhere else dictates Linux, and WSL delivers both (plus WebGPU, which is another aspect of Firefox/Linux that is perplexing and half-baked).
On both computers, X was a dramatic improvement over Wayland: proper color support (i.e. parts of the GIMP UI wouldn't turn magenta or half-transparent, you could enable HDR in Chrome), no extreme keyboard lag, all my Java binaries work without requiring cryptic flags, I didn't have to compile my own drivers from a guy named ElFarto, I could finally forget the difference between AV1 and VP9 and whichever gstreamer package lets you watch video again...
Of course, we must be realistic. Linux around 6% is very different from MacOS at 6%. In that case, they already had support from items people complain to this day about Linux not having; main examĺes are msoffice and photoshop.
Nevertheless, Linux on the desktop has never been so good. Flatpak (or even AppImages and snaps) allow me to have a "stable core" with recently updated software. Support for most hardware is much better now (looks like most vendors make sure the hardware supports Linux, even though they themselves don't announce it). Pipewire and Wayland matured to the point where you can finally stream your desktop on Wayland using Pipewire. GNOME is snappier today then it was and cleaner and more elegant; of course we still need some consistency but windows suffers from that too; and, look, it even has thumbnails on the file picker! The kernel evolved to a point where desktop is just another well supported system: MGLRU and other improvements made latency on the desktop even under memory pressure just ideal.
The environment around it also evolved a lot. Many FLOSS software are on a quality level today that mostly only professional on almost niche areas can't use Linux on the desktop. Consider Blender, Godot, Audacity, Inkscape, Firefox, OBS... these things are refined, stable, elegant and don't try to steal your attention, require periodic payments or throw ads on your face. Actually a Linux desktop user feels very sorry when they see a "standard" windows user. And I'm not even talking about proton or areas where Linux leads or is well established.
So yes, it improved. Nevertheless, I'd love to, but I don't think I'll see Linux on the desktop beat windows or MacOS. But, know one thing: I don't care. Linux on the desktop has been good enough for me for a very long time and it will only get better. I just hope someday its market share will be big enough for it to no longer be ignored by so many vendors. Watching current growth, I don't think that is too far anymore now.
I fret Everytime I update, yet I must update due to work. but I have to schedule and plan those updates so that when the whole thing takes a shit, I can spend 2 or 3 hours fixing it. Just using the built in update utility, or apt upgrade, or really anything runs about a 10-20% chance of shit breaking. (I didn't do the math)
audio experience is terrible all around, and Bluetooth and wifi use out the box, is meh at best for range and stability.
I've tried a dozen distros to see if the grass is greener somewhere else, it isn't, and I have to work, I don't have an hour every day to fiddle, I have to bill clients.
good support is almost non existent, IF you get help, it's from some greater than thou righteous asshole who suggests you just rewrite the drivers yourself and create a pull request.
while I agree it is close, there is a still a LOT of 'polish' that needs to come where I will feel confident that hitting update doesn't ruin my day, and that when I'm done with a stress filled day of work, I'll be able to boot up and play my games without having to troubleshoot for an hour first.
> I fret Everytime I update, yet I must update due to work. but I have to schedule and plan those updates so that when the whole thing takes a shit, I can spend 2 or 3 hours fixing it. Just using the built in update utility, or apt upgrade, or really anything runs about a 10-20% chance of shit breaking. (I didn't do the math)
What... are you doing? Do you have some strange use-case that might be causing this? Since 2014 I've daily driven debians, Arch, and now NixOS and updates that break my system have been exceedingly rare. I don't remember a single time Arch pacman -Syu broke my system, nor apt, and, well, NixOS doesn't count ;)
> good support is almost non existent, IF you get help, it's from some greater than thou righteous asshole who suggests you just rewrite the drivers yourself and create a pull request.
Where do you ask for help? How do you ask it? I've never received such a dickish response when I go on IRC and ask questions, no matter how stupid. I'm consistently impressed at how dedicated some community members are to helping newbies out. What you're saying is completely opposite to my experience.
If you use bleeding-edge hardware with an OS that vendors aren't specifically targeting, expect problems. I gave up trying to put Linux on random new laptops, so I've researched compatibility ahead of every purchase I've made in the last 10 years. Also, if you buy from a vendor that specializes in Linux, there will be a markup, but you will also get assurances and support.
We can pretend like that's a failure on the part of Linux, but Apple only gives you like five choices of machine, while Linux runs at least somewhat on every random computer-like object.
I've used desktop Linux since the late 90s - Redhat 6 was the first disto I used.
I have had uncountable complete system failures since then. As the parent says doing a dist-upgrade has a very non-0 chance of completely bricking the system. I've even had apt brick itself multiple times through normal upgrades.
Linux on the desktop for me over the past 25 years has been hilariously unreliable, by far the most buggy software I have used over my career.
Linux on the server is flawless and I use it all the time, but the amount of software installed compared to desktop usage is tiny, and I don't do dist upgrades on servers (would rebuilt it onto another machine).
For me after 25 years of various attempts the most reliable desktop "Linux" is actually Windows with WSL2. All hardware works great, sleep works, Bluetooth works, display scaling works, WiFi6E works etc. And I have a great Linux machine thru WSL2.
I remember often having to fix something when dist-upgrading Ubuntu when I used it on personal desktops between ~2005 and ~2014. I can't remember what kinds of issues there were, and apparently none of them entirely bricked the systems beyond repair since the same install survived throughout those years, but I remember it wasn't flawless.
But random breakage from standard updates within release is not something I remember happening on Debian, Ubuntu or modern Fedora with any frequency. (Except when I ran Debian unstable before Ubuntu, but that's a different story.) Double digit percentages, or even integer ones, do indeed sound like they are from a different planet.
I switched away from Ubuntu on the desktop over some dissatisfactions nearly a decade ago (and I never really ran Kubuntu) so I don't have much experience with that from recent years. But I've now had the same install of Fedora since 2014 that's gone through more release upgrades than I care to count, and while release upgrades still have a non-zero chance of breaking something, standard updates have been pretty much flawless.
I suspect there's something rather different about the setups of people who report no issues and that of someone who says standard updates frequently break in mainstream distros. Either there's a hardware difference or a difference in some kinds of 5% needs for the setup -- such as some specific software, specific peripherals, or specific needs. E.g. someone doing sound production might be much more likely to run into sound issues than someone whose sound needs are Netflix on USB headphones.
My last couple of personal devices have been ThinkPads that I picked in part specifically for the compatibility. The previous one with almost everything Intel worked pretty much 100% issue-free for years. I can imagine that if I instead had a random-brand consumer laptop or desktop, a wifi card with some kind of flaky half-support, proprietary NVidia drivers, some slightly niche peripherals, and the need to run some kind of a proprietary application dynamically linking with shared libraries from 2017, the story might be quite different.
Of course it might also be that some of us are just so used to dealing with fixing small issues or skirting around them that they just don't register for us any more. I think that's something that might also happen for some people -- tinkering is so natural that it doesn't feel like tinkering so tinkering doesn't exist -- and I wonder if that's part of the difference in the reported experiences.
But if someone reports a double-digit percentage of stuff breaking on a standard update, there's something else going on as well.
Having said that, my RaPi has got in a bad apt state twice now which should be fairly 'mainstream' as hardware goes (as in a lot of people running the exact same configuration).
I think you may be right with what you consider bricking too. I would consider the system bricked if a dist-upgrade succeeded (system boots ok) but apt is in a bad state after, as it can take hours to resolve that.
I think also I have a lot less tolerance for this after running Mac and Windows for so long (I've only recently started using Windows again once WSL became stable, before that it was basically just Mac). I have never had a bricked upgrade from that.
I remember having occasionally dealt with having to manually run dpkg reconfigure to install packages that had somehow ended up in a semi-installed state during an upgrade, or some other similar bad state. To someone who's not familiar with the distro's internals, and probably to some who are, that would just appear as a mysteriously broken system that no longer operates.
I guess for me bricking would be if I can't get the system working again using some kinds of relatively well-documented or known repair steps (that is, known to me or sort of generally). I'm more bothered by something not quite working right and not understanding it than I'm by a complete but repairable bricking that I understand or know how to fix.
That's almost certainly not how most people would perceive it, though.
And even though I'm probably not that averse to fixing things, I also originally moved from Debian unstable to the newly founded Ubuntu because I didn't want to deal with routinely playing sysadmin on my personal box any more.
With that said, I don't know if it's because of stricter checks being made by dnf before package operations or what, but I've never got dnf/rpm into a non-working state on modern Fedora. Despite my old affection for Debian-based distros, I admit that apt/dpkg has seemed more flakey to me after this.
After nine years, I wouldn't even know how to try doing something similar to manually running dpkg reconfigure on Fedora. If that even exists.
(edit: sentence)
I had some issues with Pulseaudio but overall they were minor, and now Pulseaudio has been swapped for something else and I don’t even know what it is… because it works well enough.
I’m not trying to say you didn’t experience problems, but your experience is the opposite of mine. I had more problems with Windows than I’ve had Linux. Windows absolutely ruined installs and disks on bad updates I more times than I can count. And the networking stack isn’t dog shit like modern MacOS (I still use Macs for work), plus unlike MacOS I can use a performant GPU to play games or do research.
This has it at 5.1%. But ChromeOS has a lot of corporate mojo making it work. It’s only partially open source. A lot of the good bits are closed. So that puts Linux at 2.91%. A rounding error. For 30 years.
Thanks for pointing how fast Linux on the desktop has been growing lately.
Google pays it to select it as the default browser, but Firefox' tiny declining market probably makes Google want to not to continue this except perhaps as a way to avoid monopoly stuff.
I'm considering writing letters to some of the sites that have the biggest issues and raising it as an issue.
Otherwise, Firefox is mostly fine. I have a few gripes with it and Moz are mismanaging their projects, but desktop Linux is hardly their #1 issue.
Or, is the problem in the complexity of the modern web browsing experience, and the walled garden created around it by current major browser developers (new standards, etc.)?
Probably the only notable attempts at building something from scratch recently is servo and ladybird. Servo was (is) an experimental platform to trial new components for Firefox is isn't a serious option for everyday use. Ladybird is primarily a hobbyist project that isn't a serious option for everyday use but has managed to implement a large part of the features of a modern working browser. The article called it "crazy", but it is impressive how far it's got so far.
Also, all of these browsers are open source with permissive licences for the bulk of their source code.
Gecko (from Netscape 6) in Firefox
And KHTML (from KDE Konqueror) -> Apple Webkit -> Google Blink in everything else.
Chrome on Linux had a random weird 3d grey border for years which was only fixed when it was announced ChromeOS was going to switch to the Linux desktop version.
Otherwise I find the experience on Windows and Mac very similar.
That's all. All the content, sidebar, etc stays stationary.
It's true that Safari does it better, but honestly not by much.
The best part, is when they do shift down, they only expose empty space.
Hardware video acceleration is fine now, and apparently has been for some time.
But e.g. Google Meet plainly refuses to blur the video call background when run on Firefox. It happily does that when running in Chromium.
And for a long time, there has been an issue where if you update Firefox with a linux package manager, you have to restart the whole browser, or the browser stops working. That isn't a problem on windows or mac (usually). And it isn't something you have to deal with with chromium on linux either.
That seems rather like a Google dark pattern where they take advantage of owning both sides of the communication channel to abuse some internal Chrome API that Firefox knows nothing about.
On ff I sync my add-ons, bookmarks and containers. What’s not to like?
Open a 1080p YouTube video and watch your processor get slammed. No, I do not want to install an unvetted third party plug-in. If nothing else, flag that the browser doesn’t support hardware acceleration when playing videos, in easy to understand terminology so I, and thousands of others don’t have to go troubleshoot exactly what is causing my computer to choke.
Hopefully then they’ll stop reinventing tabs, ruining privacy, and focus on the aforementioned bugs.
'Defining' Linux support:
>> Firefox is first and foremost a Windows browser, followed by macOS second, and Linux third. The love the Linux world has for Firefox is not reciprocated by Mozilla in the same way, and this shows in various places where issues fixed and addressed on the Windows side are ignored on the Linux side for years or longer.
80% of all revenue is from Google:
>> The giant sword of Damocles dangling above Firefox’ head are Mozilla’s really odd and lopsided revenue sources. As most of us are probably aware, Mozilla makes most of its money from a search deal with Google. Roughly 80% of Mozilla’s revenue comes from Google, who pays the browser maker to set Google Search as the default search engine.
Firefox declining usage and anti-trust risks with Google.
>> How long will this deal continue? Will it be renewed indefinitely, regardless of how much farther Firefox slides into irrelevance? Will the size of the deal drop, or will it end altogether? When will Google decide that spending hundreds of millions of dollars every year in what is essentially charity for a competitor is no longer worth it, or needed?
>> Google’s similar search deal with Apple is already facing legal scrutiny; will that scrutiny have consequences for the deal with Mozilla, too?
Solution:
>> "Linux needs a browser engine that is independent of Google (and Apple), and takes Linux seriously as a platform."
Mozilla could have done that by being a Rust consultancy which can fund the development of both Rust and Firefox without being heavily dependant on Google. That opportunity has been lost.
Realistically the likely outcome is to just break up Chrome from Google and use the Chrome engine as the standard browser engine, just like the Linux kernel is the standard kernel for all distros.
Case solved.
The Linux browser escaped out to conquer (konquer?) the world.
Konqueror was a nice browser, my fav back in the day, and it was exciting when Apple picked it to defeat IE with Safari.
(Disclaimer: I personally don't daily drive Brave, but I would consider it if I ever decided to drop Opera.)
However, maintaining a variant of Ungoogled Chromium or Vanadium targeted at desktop Linux would be feasible.
So what are you proposing then? It seems so irresponsible to complain about new-buzzword-shittification of something thousands of people contribute to and offer no contribution in return.
https://lunduke.locals.com/post/4387539/firefox-money-invest...