4 ms·

> Primarily, what I’ve learned is that Wayland is actually really well-designed. The writing is on the wall for X, and Wayland really is the future. Hmm. The f

> Primarily, what I’ve learned is that Wayland is actually really well-designed. The writing is on the wall for X, and Wayland really is the future. Hmm. The first sentence does not imply the second is true. I wonder what it would take to see widespread adoption of a new windowing system. Quite a lot, I would imagine.

nsajko 3y ago
To add insult to injury, the first sentence is false.
- Jasper_ 3y ago
  I certainly have some complaints with Wayland's design, but I'm curious what you would remark as dealbreakers here. As someone intimately familiar with both X11 and Wayland (and used to maintain stacks for both), Wayland is a real improvement along many axes.
  - a1369209993 3y ago
    > but I'm curious what you would remark as dealbreakers here. I'm not sure how to properly and (edit: semi-)concisely explain the problem except by example: How do I translate[0]: $ ssh -Y raspberry-pi-on-null-model-cable ~/bin/my-x11-c-program to Wayland, when executed from a non-ARM machine running a X server (respectively Wayland server) with a unknown compositor[2], without `my-wayland-c-program` loading any compositor-specific code or resources (and preferably being statically linked[1], although statically linking xlib is admittedly also a pain in the ass[3]). Things that are the responsibilty of the window manager, such as the title bar, close window button, and alt-space menu, need to work correctly on any compositor that supports them, without the application running any code related to those things. If it was just having to run everything on the same machine, wayland would have to have conspicuously significant advantages (of which it has roughly none) to be worth the loss in functionality. Adding on per-compositor special-casing - especially when the suggested 'fix' is apparently for me to link per-compositor shitware into my applications to allegedly abstract away that special casing - means Wayland needs to die in a fire. 0: Regarding ssh -Y: yes, the lack of security-distrust between two pieces of hardware right in front of me, that I nominally trust and control, is intentional. 1: ie `file my-wayland-c-program` prints (emphasis added): ELF 32-bit LSB executable, ARM, version 1 (SYSV), [*]statically linked, stripped[*] 2: "unknown" as in the application doesn't know, and isn't going to find out, because it doesn't matter unless the application is `wl-print-compositor-name` or similar.

nsajko 3y ago

To add insult to injury, the first sentence is false.

Jasper_ 3y ago
I certainly have some complaints with Wayland's design, but I'm curious what you would remark as dealbreakers here. As someone intimately familiar with both X11 and Wayland (and used to maintain stacks for both), Wayland is a real improvement along many axes.
- a1369209993 3y ago
  > but I'm curious what you would remark as dealbreakers here. I'm not sure how to properly and (edit: semi-)concisely explain the problem except by example: How do I translate[0]: $ ssh -Y raspberry-pi-on-null-model-cable ~/bin/my-x11-c-program to Wayland, when executed from a non-ARM machine running a X server (respectively Wayland server) with a unknown compositor[2], without `my-wayland-c-program` loading any compositor-specific code or resources (and preferably being statically linked[1], although statically linking xlib is admittedly also a pain in the ass[3]). Things that are the responsibilty of the window manager, such as the title bar, close window button, and alt-space menu, need to work correctly on any compositor that supports them, without the application running any code related to those things. If it was just having to run everything on the same machine, wayland would have to have conspicuously significant advantages (of which it has roughly none) to be worth the loss in functionality. Adding on per-compositor special-casing - especially when the suggested 'fix' is apparently for me to link per-compositor shitware into my applications to allegedly abstract away that special casing - means Wayland needs to die in a fire. 0: Regarding ssh -Y: yes, the lack of security-distrust between two pieces of hardware right in front of me, that I nominally trust and control, is intentional. 1: ie `file my-wayland-c-program` prints (emphasis added): ELF 32-bit LSB executable, ARM, version 1 (SYSV), [*]statically linked, stripped[*] 2: "unknown" as in the application doesn't know, and isn't going to find out, because it doesn't matter unless the application is `wl-print-compositor-name` or similar.