Hacker News Reader

Comment by yjftsjthsd-h | Hacker News Reader

> In the last release post we hinted at how Xarcan can be used to keep your own window manager, letting Arcan act as a display driver — as well as a security, control and configuration plane that selectively merge in arcan native clients with options for how, or if, X clients gets to see inputs and clipboard action.

Alright, I'm probably sold. A couple of Wayland's greatest faults are that 1. they enthusiastically broke everything that X provided except for displaying windows on screen, and 2. after 16 years so far, they haven't actually provided replacements. So anything that lets us improve on the X server while keeping backwards compatibility is a huge leap forward; I just want to see if the compat is good enough for accessibility and automation tools[0] to work smoothly.

[0] This is 1 category, not 2.

View on HN

If you read through the posts, Arcan is probably way ahead of Wayland with accessibility features. A LOT of thought has went into nearly every aspect of its architecture.

A slight nuance: Yes, I absolutely agree that Arcan is winning on a11y features overall; I'll go stronger and say that it doesn't just exceed Wayland, I think it's better than X was. What I meant specifically is that I want to see if it's backwards compatible with at least some of the tools I use on X, because it looks like (especially from this post) Arcan might be able to hand existing tools an X-compatible interface (doing actual access control and everything behind the scenes but that's all invisible to the tools using this interface) and let me keep using what I already have without having to (re)write anything new. As an example: On X, I can use `xdotool type --window $WINDOW_ID "sometext"` to type into a window even if it's not active. On Wayland, your best hope is that you happen to be using a compositor that's implemented some nonstandard API to do that. My hope is that Arcan implements enough of X that you can make it work, albeit probably with some explicit permission-granting.

Of course, Arcan looks good enough, and actually exposes enough API surface, that if it's not compatible with my existing tools then yes I probably will write whatever I need over its native APIs and just run with that. But it'd be nice if I didn't even need to do that:)

I see what your meaning now, thanks for making it clear. I haven't used Arcan in a while, but from what I remember from my (admittedly little) use and architecture reading, it might be in the best spot to get what you want. Lua bindings are pervasive and should in theory get close to what you want, but more research and work is needed to realise that.

But yes, I really feel bad that after 16(!) years, wayland is still a shitshow in all the cases of not drawing windows on the display, and I say this as someone who has been using Wayland daily for half a decade now, and lived through things being even shittier. I just silently keep wishing Arcan gets there as a universal drop-in replacement.

I can't find the link now, but there are two ways (I know of) for getting the xdotool way to work. One way is the Durden 'control' socket and a standalone tool 'arcan-cfgfs /path/to/.arcan/appl-out/durden/control /my/mountpoint' that is a FUSE driver. You can mount the desktop as a file system and automate everything like that: echo "typethis" > /my/mountpoint/windows/all/text/input". Another is to expose that more generally with hookscripts (force-injected Lua scripts that override or add features to the WM you use, arcan -Hhooks/something.lua).

Reply on news.ycombinator.com