Future of GTK2 and Pale Moon

[athenian200]

I kindly caution against this approach, simply because I fall in the unhappy, neutral ground between the two classes you discuss. Because I know you are also autistic, I can speak forthrightly: The kind of involved, often unpredictable technical work which building software (or much troubleshooting) involves is the kind which tends to provoke panic attacks in me. Although I do not imagine I would be personally affected by your decision here, it is easy for me to conceive of somebody else like me in this respect. In our case, being a responsible adult means recognising we need to protect our emotional health against hazards like those uncoöperative IT can provoke.

Then... in all honesty, my response to that would be that you should accept your limitations and use what is provided, even if what we decide we can reasonably provide winds up not being to your liking. If you can't handle the work, I'm sorry, but that doesn't give you the right to burden someone else because of your preferences. We provide binaries precisely because we know some users can't just do it themselves, but there are limits to that. That's the situation a lot of people are in. Maybe their computers are too old to build a UXP application, maybe it stresses them out, maybe they suffer from cognitive decline and can't do it now even if they would have been able to do it years ago. There are going to be a lot of sad stories in there... stories that seem to justify our continued intervention and going well out of our way to provide GTK2 for hapless people who just can't do it themselves. And those stories will not go away no matter how hard supporting GTK2 becomes for us. At some point, we have to look out for ourselves. Hopefully that doesn't seem insensitive?

In any case, I believe that outright depreciating NPAPI for Linux, even if it is due to underlying GTK2 changes beyond your scope, would be unwise because of the inevitable backlash. It would be far too easy for someone naïve or malicious (and you know there are many of these out there) to portray Pale Moon as reneging on its promise to preserve NPAPI support, even if this is no more in your hands than Flash’s incompatibility with younger Glibc versions if run with Nvidia hardware. Better in the worst case, I think, to keep nominal NPAPI support with the proviso for the user that most distros have torn the ground out from under his feet.

That's exactly why I'm investigating whether there's a way to avoid GTK2 without getting rid of NPAPI on Linux entirely. Getting rid of it on Linux is a worst-case scenario, and I want you to rest assured I'm looking into the problem so that we can support it in some form on Linux without being tied to that toolkit. There are usually ways to solve problems when it comes to computer programs... not always easy ways, not always obvious ways, but usually something can be done.

But yes, I'm fully aware that NPAPI is one of the killer features for UXP. It's not something I'm eager to give up on, myself. Like I said, I'm not very happy with the state of our Linux code and do in fact feel like Linux is the platform on which we are furthest from keeping our promises to the userbase on these things. We don't need those hypothetical people you're talking about to guilt-trip us, believe me, we are aware of the mismatch between what we set out to build, and how our code on Linux increasingly looks.

Maybe you and a few others see more people using Linux as a victory, but for me? All I'm thinking is... great, more previously happy Windows users of UXP will see how much worse the Linux version is than the Windows version and quite possibly join the GTK2 crowd in holding our feet to the fire in a way that increasingly burns them in an unbearable way, rather than motivating us to move. I'm glad to hear some Linux users are more fair-minded (their support in this thread has meant a lot), but I'm dreading some percentage of Windows users migrating to Linux and falling right into the GTK2 camp with all the other stubborn non-tech savvy types and making our lives harder. They're historically Windows users, which means they are used to first-class service and backwards compatibility, but now they're on Linux, a platform that pushes "upgrade or die." People who are Unix people? They're usually fair-minded and understand the limitations, that it's up to the distro. But dealing with Windows people that have "gone Linux" to get away from Windows 11? That's what I really don't look forwards to...

[Mæstro]

Then... in all honesty, my response to that would be that you should accept your limitations and use what is provided, even if what we decide we can reasonably provide winds up not being to your liking.

In the bit of my post quoting Andy Prough, I sketched how I would treat the moot case of Flash failing in Linux if it were to face me. The methods I described are ways I could handle within my limitations. Other possibilities could include, off the top of my head, hiring somebody to provide NPAPI-compatible builds for me, or if I really felt out of options and forced to it, sticking with the last NPAPI-compatible versions both of Linux and of Pale Moon for Linux. In any case, such a failure would obviously not be your fault in any way.

That's the situation a lot of people are in. Maybe their computers are too old to build a UXP application, maybe it stresses them out, maybe they suffer from cognitive decline and can't do it now even if they would have been able to do it years ago. There are going to be a lot of sad stories in there...

You understand. You really do. Ensuring that you do was my purpose in describing my own position. I should likewise apologise if I seemed to wish to make you feel guilty in my earlier post. I knew my phrasing was poor, but could not think of any better.

At some point, we have to look out for ourselves. Hopefully that doesn't seem insensitive?

I do not want you to go mad, either. I understand.

Maybe you and a few others see more people using Linux as a victory, but for me? All I'm thinking is... great, more previously happy Windows users of UXP will see how much worse the Linux version is… But dealing with Windows people that have "gone Linux" to get away from Windows 11? That's what I really don't look forwards to...

I admit I am awkwardly near that class. A close friend, the one who just changed over, is such a Windows 11 refugee. In my case, Linux has permitted me to keep away from Windows 10 and 11 in turn, so I have always been able to look on them aloof from afar. This line is most revealing:

They're historically Windows users, which means they are used to first-class service and backwards compatibility, but now they're on Linux, a platform that pushes "upgrade or die."

Speaking, of course, as a layman, my perspective has been upside-down. Windows 10 was, and Windows 11 is, unfit for my use (among other reasons) because of their own ‘upgrade or die’ mentality, chiefly due to forced updates. The long-term releases of at least some Linux distributions, as long as the better Windows versions, led me to think of Linux as the more patient of the two, as long as one kept away from bleeding-edge distros like Arch. Over five years of continuous Linux use later, I can feel more often how many developers tend to neglect even slightly older editions of the same distro, even if they remain in long-term support. I now see the error of my first impression.

Since Windows remains intolerable and shows no signs of ever getting better, what is there for me in years to come? In the last few days, I have been thinking to myself that I should purchase a Raspberry Pi 400 or two to keep on hand as spare computers in the event this one fails. Assuming I am running Debian 13 on it, that could carry me securely to 2035. What then? So much depends on the precise conditions several years hence which are impossible to predict. Maintaining the software I like against an ever more hostile development mainstream is something I can only realistically address day by day, month by month, year by year, as problems present themselves. Your efforts help make the way easier for me, and you have my deepest thanks for this.

[athenian200]

You understand. You really do. Ensuring that you do was my purpose in describing my own position. I should likewise apologise if I seemed to wish to make you feel guilty in my earlier post. I knew my phrasing was poor, but could not think of any better.

Yeah, it's all good. You didn't say anything wrong.

Speaking, of course, as a layman, my perspective has been upside-down. Windows 10 was, and Windows 11 is, unfit for my use (among other reasons) because of their own ‘upgrade or die’ mentality, chiefly due to forced updates. The long-term releases of at least some Linux distributions, as long as the better Windows versions, led me to think of Linux as the more patient of the two, as long as one kept away from bleeding-edge distros like Arch. Over five years of continuous Linux use later, I can feel more often how many developers tend to neglect even slightly older editions of the same distro, even if they remain in long-term support. I now see the error of my first impression.

Since Windows remains intolerable and shows no signs of ever getting better, what is there for me in years to come? In the last few days, I have been thinking to myself that I should purchase a Raspberry Pi 400 or two to keep on hand as spare computers in the event this one fails. Assuming I am running Debian 13 on it, that could carry me securely to 2035. What then? So much depends on the precise conditions several years hence which are impossible to predict. Maintaining the software I like against an ever more hostile development mainstream is something I can only realistically address day by day, month by month, year by year, as problems present themselves. Your efforts help make the way easier for me, and you have my deepest thanks for this.

Yeah, that is worth expounding on. My perspective is this... Windows has, even at the Windows 10 and Windows 11 level, been much better at supporting older binaries and keeping a stable ABI for things like plugins compared with Linux. So for developers, it's a more forgiving platform. You don't have to do nearly as much work to support Windows 10 if you supported Windows 7, or nearly as much work to support Windows 11 if you supported 10. Windows lately has a nasty habit of breaking the user experience and pushing updates (though I hear they're considering bringing back the way to pause them after the amount of backlash recently), but is still a lot more forgiving of older binaries and providing ways to get code with older assumptions to compile cleanly. Whether you love or hate Windows 11, it can still be used to compile Pale Moon and still runs all our tooling. So it's stable for the developer, and still runs most old programs, but manages to piss off users with an ugly UI and forced updates in recent versions.

Linux is... almost the reverse, tries to avoid breaking userspace, keep things looking the same for users across versions (at least on the surface), but tends to break things for developers under the hood. Something like MATE might even fool some people into thinking not much has changed since GNOME 2 unless you wanted it to, that you have choice and freedom, and everything is still fine and nothing has changed in a big way. When in reality, MATE is using GTK3 (rather than GTK2) in a very careful way to recreate the GNOME 2 experience, systemd stuff is symlinked to traditional sysvinit commands, and really everything that looks familiar to someone on the surface may well be powered by something totally new under the hood. Linux is... weirdly good at making things feel like they haven't changed as much as they have for users who know a little but aren't experts, while forcing to developers to get on the upgrade treadmill or watch their work collapse beneath the waves. Much like the web with Chromium, I'd actually compare the Linux churn on the platform side to the Chromium churn on the web side. The website looks the same and does the same thing as it always did in Chrome or maybe even Firefox, but it pulled in a newer version of a framework that uses new mechanisms and now it breaks in Pale Moon. Linux feels like it's driven by the same mentality. That's where the web/Chrome analogy comes in.

Since Windows remains intolerable and shows no signs of ever getting better, what is there for me in years to come? In the last few days, I have been thinking to myself that I should purchase a Raspberry Pi 400 or two to keep on hand as spare computers in the event this one fails. Assuming I am running Debian 13 on it, that could carry me securely to 2035. What then? So much depends on the precise conditions several years hence which are impossible to predict. Maintaining the software I like against an ever more hostile development mainstream is something I can only realistically address day by day, month by month, year by year, as problems present themselves. Your efforts help make the way easier for me, and you have my deepest thanks for this.

Yeah, that might be your best option for now. Debian's ELTS offering is apparently nothing to sneeze at, and I just checked... that will go to 2035, but you do have to pay. Aside from that, my instinctive recommendation would be possibly RHEL 8, a free RHEL rebuild (like Oracle Linux 8), or some version of Ubuntu LTS. RHEL gives you 10 years of support, and Ubuntu LTS might offer more (though I haven't looked into that yet), plus it might feel more familiar as a Debian user. Supposedly RHEL will even go to 14 years, but it's insanely expensive.

Let me put it this way... we like to build our applications on enterprise versions of Linux for many reasons. One of them is that we get a frozen stack/ABI for a lot longer and things don't change out from under us as fast. The other is that binaries we produce on such systems tend to work across a wide range of Linux versions, which isn't guaranteed if you build binaries on any old Linux distro. It's not necessarily because we like enterprise versions better, it's because that seems to be the only way to get a time frame approaching sanity in a lot of cases as far as deprecation windows for important software packages. If paying people is on the table, it's much easier to find someone willing to offer extended security updates for RHEL or one of its forks, or maybe Ubuntu LTS, than for just about anything else. Those are the versions the Linux long-term support "industry" has standardized on, for better or worse.

[andyprough]

RHEL gives you 10 years of support, and Ubuntu LTS might offer more (though I haven't looked into that yet)

Ubuntu LTS is the better deal for end users - 15 years of security updates for free if you sign up for a free Ubuntu Pro account which covers 5 systems. Very hard to beat, and the 3rd party software support for Ubuntu is stronger.

I personally dislike Ubuntu for a lot of different reasons, but that deal is too good to pass up, so I've got my free Ubuntu Pro setup and my own 15-year Ubuntu LTS on a VM.

[athenian200]

Ubuntu LTS is the better deal for end users - 15 years of security updates for free if you sign up for a free Ubuntu Pro account which covers 5 systems. Very hard to beat, and the 3rd party software support for Ubuntu is stronger.

I personally dislike Ubuntu for a lot of different reasons, but that deal is too good to pass up, so I've got my free Ubuntu Pro setup and my own 15-year Ubuntu LTS on a VM.

That's a good tip for users, I think we might actually want to recommend that. I just happen to know also from inspecting readelf output and binary strings, that the version of Ubuntu the last version of Adobe Flash Player was compiled on... was quite likely Ubuntu 14.04. So that's another point in its favor.

[Mæstro]

Yeah, that might be your best option for now. Debian's ELTS offering is apparently nothing to sneeze at, and I just checked... that will go to 2035, but you do have to pay.

Two years ago, as Debian 10’s LTS was about to lapse, I wrote to Freexian explaining my situation and enquiring about costs for personal use. A sales representative contacted me to tell me that Freexian would waive all charge for me personally. For their understanding and kindness (it would have been about €600 a year), I am most grateful.

Ubuntu LTS would very much interest me if not for my attitudes towards Canonical and my impression that it would require a good deal of purging to remove any trace of the Snap Store, Flathub etc from my system if I were to use it. This is, of course, personal, and should not alter your general advice. I should keep Red Hat in mind as a future possibility also, but my impression is that developers give it less attention than Debian and its kin.

You would be in a better position as a developer to confirm or deny my speculations, but I would guess that Linux’s own pressure to use newer versions of distributions is partly also because of Linux’s small usage share, which is further fractured among so many distributions. Preparing three Debian editions in parallel is work enough, but adding an Arch version, Fedora etc etc adds more work than any small-scale developer can reasonably be expected to bear. Supporting more new distributions surely appeals more to developers than many older versions of the same one, and so the older versions are sacrificed.

Comparing Linux’s lot witrh Mac’s would be instructive, not only because both are Unixlike, but because Mac’s usage share twenty years ago resembled Linux’s now, and I am old enough to remember everything said about software not supporting Mac. Since individual Mac OS versions are officially supported only a few years, yet the official Pale Moon build for Linux supports Lion and up, I am left to think that Mac has been more conservative across versions than Linux is. If the latest Pale Moon update requiring FreeBSD 14 is any indication, FreeBSD is more like Linux in this respect. I would have no way to gauge how volatile the other BSD are.

[andyprough]

Ubuntu LTS would very much interest me if not for my attitudes towards Canonical and my impression that it would require a good deal of purging to remove any trace of the Snap Store, Flathub etc from my system if I were to use it.

I don't actually use Ubuntu for my desktop. I really strongly dislike Gnome for any productivity tasks. But as a free, super longterm support distro for setting up a build environment in a VM and leaving it as-is for a decade or longer? For sure I'm going to want to use it. And if I wanted a super longterm stable VM for doing my NPAPI stuff inside - I absolutely would use it. Set it and forget it.

[jarsealer]

Ubuntu LTS would very much interest me if not for my attitudes towards Canonical and my impression that it would require a good deal of purging to remove any trace of the Snap Store, Flathub etc from my system if I were to use it.

If I remember correctly, 24.04 LTS, which will still be supported for a while, was the last release before snap could break some things if you removed it from your system.

[Donieck69]

https://git.devuan.org/Daemonratte/gtk2-ng gtk2 fork for modern and safe gtk2