Ubuntu's Mir Finally Supports Drag & Drop

[Mateus Felipe]

is there any rational reason why would you want to replace it?

Yes. It's old and obsolete. Stableness isn't quality.

[leipero]

Yes. It's old and obsolete. Stableness isn't quality.

For me, those are not rational reasons, wheel is also old, very old, that doesn't make it obsolete tho, X is still functional, can't say teh same for otehr protocols .

[GI_Jack]

It looks like Wayland and Mir are both in perpetual alpha/beta.

oooo, biiig diffrence here. Wayland is in late beta, pre-release stage. GNOME has already switched over to wayland default, and there are many other ports in the works.

I am posting this from a gnome/wayland arch gnu/linux system. Everything "just works", and I didn't notice I was running wayland until I started looking for the X server to find that the only X server running was XWayland.

Everything just works with no noticeable differences, including X apps using XWayland. Mabey not all drivers, but intel ftw

[Iksf]

there are security concerns with X due to its huge codebase, questionable code maintainance, redundant and unused featutes etc. Things like composited desktops and other 3d things particularly are pretty hacked in rather than an optimal design. Theres some issue with sandboxing as well. Wayland will eventually be more flexible, lighter, faster, more stable and have less arcane hacks to fix niche problems.

[Iksf]

As for X getting hate, think its more a running joke atm

[panda84]

I often read comments about how X11 is old, but besides that, is there any rational reason why would you want to replace it? Wayland is terrible at current stage, it may fit some (or even majority) use case scenarios, but it is not even close to X in terms of funcionality. MiR, we know it lags behind Wayland even.

I am not against introducing new (and potentially better) protocols, it's great, but why so much "hate" towards X? It is without question superior to any protocol, otherwise it wouldn't be used on de-facto every GNU/Linux distribution.

For my use case scenario, none of the new procotols could reach even basic functionality terms, let alone advanced things. For example, wayland relies on EDID information, when you are in use case scenario where EDID information is wrong, you have to jump trough tons of hoops, build your own EDID binary, load it in KMS, and pray to the Gods of Egypt it will work. How is that better than simple modeline in X? It can't be, and it isn't. I'm not even touching root privs...

TL;DR: X is good, and it is still far superior to any otehr protocol.

If you are actually interested in the answer, it's all in here:

https://youtu.be/RIctzAQOe44

If you meant to whine "just because", skip it, it's quite long and in depth.

[TheBlackCat]

I often read comments about how X11 is old, but besides that, is there any rational reason why would you want to replace it?

A lot:

1. X11 is inherently insecure. Every application can read everything from every other application. And there is fundamentally no way to make a secure screen locker.
2. X11 is full of legacy cruft that nobody uses anymore. This means it is more complicated to program for, much, much harder to maintain and improve, requires much more resources than it needs, and doing anything requires going through a ton of pointless intermediate steps.
3. Due to its design, certain basic things are fundamentally impossible to do reliably, like keeping window borders and window contents lined up, keep track of child windows, keep track of what application owns what windows, and keeping video elements inside a larger UI positioned correctly.
4. Due to the way different parts of X11 work together, one application can freeze the entire desktop.
5. A lot of components that are needed were tacked-on and cobbled together over time, resulting in them working poorl. The compositor, which is the way pretty much all modern desktop environments work (on Linux or otherwise), is a tacked-on extra in X11. The multi-monitor management is pretty flaky. The network display system is terrible for modern widget toolkits. The list goes on and on.

These problems have all been fixed in Wayland.

Wayland is terrible at current stage, it may fit some (or even majority) use case scenarios, but it is not even close to X in terms of funcionality.

That is by design. X11 does a ton of things that are absolutely useless or downright harmful for a modern desktop. Wayland gets rid of all that legacy stuff in order to focus only on what modern desktop really need. The display server shouldn't have its own widget toolkit, it shouldn't have its own font handling, it shouldn't do its own input handling, it shouldn't have its own window manager, it shouldn't have its own IPC system, it shouldn't handle direct hardware access where well-established kernel APIs exist, etc. X11 has all of these, Wayland has none of them. Where those things are needed, they are provided by dedicated tools designed for that particular purpose, and kept intentionally separate from Wayland to improve maintainability, ease transitions (many work under X11 too), and allow us to avoid having to replace the entire stack just to replace one small component like we are with X11.

For example, wayland relies on EDID information, when you are in use case scenario where EDID information is wrong, you have to jump trough tons of hoops, build your own EDID binary, load it in KMS, and pray to the Gods of Egypt it will work.

No, it doesn't. A particular DE's screen configuration dialog may rely on that, but there is nothing in Wayland that prevents a compositor from overriding display information. KDE's kwin has kscreen for handling this identically under X11 and Wayland, although according to its designers Wayland's multi-screen handling is much better than X11s.

How is that better than simple modeline in X?

You mean like this?

[pracedru]

I often read comments about how X11 is old, but besides that, is there any rational reason why would you want to replace it?

There are a few good reasons.

- X11 is not very modular in design. As such many things that arent needed, for say embedded devices, are in the way.
- The compositor part of x11 works but x11 was designed to do most of the graphics in the X11 server not in the client application. For this reason X11 doesn't perform as well as Apples Quartz or MS's Compositing Window Manager, which is very noticable on low power devices.

These things could have been solved in X11, but what is the harm in making something new if the old is still compatible?
Is it not a good thing to let old software design be as they are, and start on a fresh clean slate, when new concepts come along?
Could one not argue that X11 has been in effect destroyed a little by trying to implement stuff that it wasn't originally designed for in it?

Please also notice that X11 won't be scrapped now that wayland is here. All the stuff you love about X11 will be supported, propably as long as any one has an X11 application that they need.

[leipero]

You mean like this?

That doesn't work, it is not universal setting like X modelines. Run GNOME 3 on wayland and try to use non-EDID mode.

[TheBlackCat]

That doesn't work, it is not universal setting like X modelines. Run GNOME 3 on wayland and try to use non-EDID mode.

No, it isn't universal, nor is it meant to be. The display server shouldn't be handling configuration like that, that is the sort of the desktop environment should be (and is) in charge of. Dealing with all the different ways X11 could be configured, and conflicts between them, is a major problem.

My point was that your statement about Wayland requiring you to "build your own EDID binary" is wrong. Wayland provides the APIs necessary to change the display settings in an even better and more reliable way than X11, and it looks like Weston, KDE Plasma, and GNOME now all provide interfaces to configure them.

What about the rest of what I wrote? You responded to 4 words out of the several paragraphs I wrote.