Notices by lhp (lhp@mastodon.social)

@tk Games are supposed to still work, yes.

In practice it will depend on the game: On X there are like five different ways to implement each feature a game may need. Some of those may be a bit hacky and won't work smoothly on a Wayland desktop.

From what I understand though, valve has put in some effort to make it as smooth as possible.

I myself don't play games anymore, so I can't share any first-hand experiences.

But you can install a wayland desktop alongside your current and just try.

@tk those are two different concepts.

An application can't arbitrarily become drm master and control the output (there are some exceptions for VR stuff, but it's not common). This is the reason why tabbing out of fullscreen games is seemless on linux but is a crashy mess on windows.

But you can of course still do accelerated rendering with OpenGL. If done right, even zero-copy with direct scanout and proper frame timing.

Going through XWayland will add a few roundtrips to that.

@praetor @amszmidt you should never eval your init manually, it is loaded automatically on startup.

I suspect you set some settings in your init which don't affect the frame spawned by emacsclient. And when you then re-eval your init manually, those settings suddenly work. Is that right?

@jwz say what you want about enterprise, but at least when it was bad it was "more chliché and more boring than average TNG episode"-bad, not "we don't understand the franchise, so let's just make a generic modern action sci-fi show, only the colour blue is allowed, everything must be gritty and grimdark"-bad

@anemofilia I do really like guile, but I also like systemd, so I am not really the target audience for Guix, unfortunately :/

@ifreund Nice, thanks!

I am right now trying to migrate my setup to use more flatpaks, especially for anything internet-facing. I really like the qubes OS idea, but my main setup is too integrated to be split into VMs. Having at least browser, email and IRC sandboxed is the next best thing I think. Maybe I'll even forbid programs running in the host OS from talking to the internet eventually.

A live REPL attached to a running window manager. Eventuallly you'll be able to change window management behaviour without even having to restart the window manager.

#riverwm

This is:

* a C library doing the Wayland bits and low-level logic
* a guile Scheme module wrapping the C library
* a guile Scheme script implementing high-level logic

It's all based on hooks, so eventually the WM could just load an init.scm from the your config where your can overwrite whatever handlers / functions you want. Kinda like how you can redefine emacs functionality.

And you could also just use the library itself and wrap it in whatever language you want (as long as it has a C FFI).

connecting a WM written in guile scheme to river's rwm branch 👀

(prototype, doesn't actually do anything yet)

#riverwm

New blog post: Painless multi-account emailing with imaps and smtp in gnus in #emacs

The "painless" is of course a bit of a joke; I hope by sharing how I made it work other people are spared a few hours of config-fiddling. Once you know how it works, it is actually somewhat simple.

https://leon_plickat.srht.site/writing/painless-mail-imaps-smtp-gnus-emacs/article.html

whats this? a wayland screansaver of course!

@monnier I wasn't aware that existed when I started this project.

Initially starting out as a fun hack in between talks at a conference, iwd.el is now - thanks to a contributor - reasonably functional and can be used to connect to wireless networks via #emacs (if you are using iwd standalone).

https://git.sr.ht/~leon_plickat/iwd-el

@prahou I showed you my ~/.config/emacs/init.el, please respond

@dnkl however the protocol isn't really documented. There is an assuan doc for the wire protocol and there is one for the pinentry protocol, but they disagree with each other and with reality. Also sometimes they use "character" when it really should be "byte". Not that anything actually checks line lenghts. Pinentry protocol requires lines lengths which are technically illegal with assuan, but no one seems to care there either...

@dnkl when I implemented the pinentry protocol, first I copied the messages sent by the "official" pinentry implementations bit-for-bit. It did not work. Then I tried following the protocol spec, which did end up working, kinda. However this means the official ones are doing it wrong, but somehow still work in a way I could not reproduce nor did I ever figure out why they are allowed to do it wrong while my program gets killed by gpg-agent.

@dnkl the protocol actually has a way to set $DISPLAY without needing a command arg. In fact, there are two different ways to set various variables.

Also there is a handy error code to indicate your implementation doesn't support some feature. But don't think about using it, because gpg-agent will just abort. Instead always say OK and silently drop it.

gpg is the best argument against hand-rolling IPC.

Just merged patches updating wayprompt, a #wayland native pinentry implementation using the layer-shell, to #zig 0.13. It was stuck on 0.11 for a while, so this is - at the very least - a nice QoL improvement for users compiling from source.

Thanks @uncomfyhalomacro for doing the work! I only needed to change a few minor things. And thanks #emacs and magit for making the workflow of pulling patches from another remote, then reviewing and amending them, very comfortable.

@aral @nekohayo @nah @fvsch @sonny @matt You literally were just told - multiple times - that people know the problem and care about it and that this change is not even in the next version precisely because people are aware of this and other problems. You were told there are active and funded efforts to imprve the linux accessibility stack. Isn't that exactly what you want? Your last message in this thread feels very dishonest. Maybe there is a miscommunication here?

@lispi314 @dushman @astrid It would be nicer though if it used scheme instead (really sad guile emacs died), used more system libraries (slowly getting there) and shipped less lisp code by default. The last point is not because "bloat", but because if you want non-standard behaviour you have to override a lot of stuff which often feals janky. F.e. evil just does not feel good and it's a popular package. And if you want less popular things like kakoune-like keybinds and the experience is just sad