The elephant in the room with GTK2.

I've worked for a very long time on a project that still supports both GTK2 and GTK3 versions, and may well never be able to move to GTK4. I have had to interact long-term with users who are relentless in demanding that we continue to provide a GTK2 version and view even GTK3 with suspicion. I think I've come to a conclusion about why GTK2 is dying so slowly, and also why I think the situation is unlikely to improve.

Most of the people who love GTK2 are those who use a keyboard and a mouse on a PC with no touchscreen. Newer versions of GNOME (and hence GTK because the GNOME team does most of the work on it) are not really improvements on the previous ones for the needs of users who still use computers essentially the same way they did around 2010 or so, and if anything are alienating to those users because they are designed primarily around touchscreen-centric use cases, and possibly the UI expectations of users who grew up with Android and iOS before they ever touched a standard x86-64 PC.

It's essentially the Linux equivalent of the Windows 8 problem, where Microsoft put everything they had into adapting to touchscreens and the mobile ecosystem, neglecting traditional desktop users and expecting them to go along with the program while they attract a different kind of user and grow their brand. This resulted in many Windows users sticking with Windows 7 as long as they could. GTK2 became for a long time the Linux equivalent of Win32 (essentially taking on a niche formerly occupied by Motif in the Unix world), probably much to the chagrin of GNOME's developers and fans of other toolkits, and that situation probably isn't sustainable in the long-term. It also connects with the frustrations of those annoyed by mobile-first (or as I call it, desktop-last) web design.

Now, from the perspective of most open-source developers, working on something like GTK2, or a modern equivalent, isn't interesting because it's seen as a problem that was solved long ago, not a new challenge, and plus there's not a lot of money in it. What most of those users actually want is essentially boring security updates and tooling updates to something that's still working fine for them, not fundamentally new paradigms or a big change. They want essentially the same kind of stability a lot of big enterprises want, only they aren't in a position to pay a lot of money for it.

So there's fundamentally a mismatch between what the average open-source developer is interested in working on and where the money is going on the one hand... compared against what a lot of traditional non-touchscreen desktop users actually want Linux to be. Since most of these users have neither the coding skills nor the money to change the course of things, the river of money and coding skill will likely eventually drag those users kicking and screaming in a direction they loathe, but they will go as slowly as possible and fight like they're cornered, possibly winding up on things like LTS releases of Linux distributions, or forks of forks maintained by a very small number of people.

I think in the end, we'll see all kinds of coping mechanisms. Some projects will bring GTK2 in-tree out of desperation and keep trying to make it work in some form, evolving it in project-specific directions. Some will move to GTK3 and use that into the ground until GTK5, some may switch to different toolkits entirely.

I don't think the problem will get better, because as far as I know, there's just not enough talented developers interested in writing the modern answer to GNOME 2 and GTK for Wayland and providing a path for those who just don't want modern, touch-centric desktop paradigms and instead prefer a traditional experience. I think if something like that existed, people might be willing to put in the work to migrate, but as it stands, they prefer to just stick with the old thing as long as they can, because they don't actually need or want anything newer GTK provides, they're just stuck because security updates and tooling issues are forcing their hands and requiring them to "be practical" and accept something that, from their perspective, is worse in every way other than simply being actively developed and supported. In the end, it's just going to be a lot of very unhappy users, resisting what's new and supported as long as they can, and sooner or later having to struggle to make something work for them that was never designed around their needs.

I genuinely feel bad for everyone involved... the upstreams who just want to get on with developing the future and feel bogged down by people asking for legacy support. The end-users who just want to have cool themes and use their desktop with a keyboard and mouse like always, and don't understand why everything is locked down, limited, less customizable, and designed for touchscreens now. Not to mention the distro maintainers caught in the middle, who know upstream doesn't have their back, they can't maintain these packages themselves, and that the longer they wait to disappoint the users, the more work they take on, and the worse the bitrotting and security issues are going to be. And I honestly see this more as a slow-moving train wreck than as anything that will lead to a happy ending for most people involved.