Hi all. Some commercial projects I’m involved in use broadway as a fallback option if we can’t get gtk and our app installed on a desktop. I’d very much like to fill in some missing functionality - in particular copy+paste ( ie between apps running in broadway, and apps external to the browser ). I brought this up, a few years ago, and at that time, there was a suggestion to approach organisations linked on https://www.gtk.org/support.php … which I’ve done, and haven’t even gotten a single response, which is a bit unsettling.
As some people might know, I’ve personally done some work around securing broadway apps and handling launching apps from a regular html login page and proxying connections: https://github.com/dankasak/broadway_proxy … I’ve got another set of patches for this project on the way to make initial setup much nicer, and add https termination, by the way.
It’s heartening to see that broadway has been kept alive for gtk 4. There aren’t many other features to go, and we’ll have a 1st-class delivery method for gtk apps.
So now I’d like to restart the discussion. Ideally, some gtk devs would find it worthwhile to tackle this if offered sufficient funding. Otherwise, I guess I can try some of the ‘rent a coder’ type sites, though I assume the quality would be worse, and there’s a risk that as a result, patches would not be accepted.
I’d appreciate it if gtk devs could respond and share their thoughts.
If there are others who would also like to see this implemented, and would be interested in being a part of a crowd-funding effort, also, please respond.
We only support windowing system backends if people contribute to them; at most, we can guarantee that a backend keeps working within a baseline level, but feature work (and bug fixing) only happen if somebody relying on that backend actually puts work on that. That applies to any backend; the reason why the Wayland backend is currently favoured is because most of the GTK developers actually use a Wayland-based system, for instance.
The Broadway backend is very special, in the sense that it’s kind of a science experiment; it’s not entirely unexpected that you won’t find any support for it—unless you’re willing to pay enough engineering time. There are only a couple of people who understand it, and they are otherwise busy with other pieces of infrastructure.
Any patch would have to go through the same coding review and iteration, regardless of provenance. The only thing that asking an existing GTK developer, or any existing consultancy with GTK development experience, buys you is that they would already know the cost of the iteration needed to get a contribution merged upstream; we don’t really play favourites, especially if people disappear and we get to maintain their code for the next 5-10 years.
