Canonical to drop x-org for "Wayland".

[woodsmoke]

Since I have no clue I will leave it to the people smarter than myself to comment.

This is a November 2010 article so it may have already been discussed and I missed it while taking a nap.

http://www.thevarguy.com/2010/11/08/...-with-wayland/

woodsmoke

[james147]

I hope this happens, wayland has allot of promise... I just hope ubuntu don't do what they have done with all the other promising techs that have been coming out over the years (pulseaudio for one) and adopting it before it is ready or prime time. From what I remember the developers of wayland said they don't think its going to be ready at least until late 2012, early 2013.

From what I know, kde has made the most progress to being able to work on wayland (thanks to qt building support for it early), with kde4.8 having at least some support for it, though no where near ready to switch yet... My guess is it won't be ready for the next release, but hopefully the release after... though I haven't been following the development lately its mostly just a wild guess .

[woodsmoke]

thankeee

woodsmoke

[GreyGeek]

I ran this on my Precise installation:

$ locate wayland
/usr/share/doc/mesa-common-dev/WL_bind_wayland_display.spec
jerry@jerry-Aspire-7739:~$ cat /usr/share/doc/mesa-common-dev/WL_bind_wayland_display.spec 
Name

    WL_bind_wayland_display

Name Strings

    EGL_WL_bind_wayland_display

Contact

    Kristian Høgsberg <krh@bitplanet.net>
    Benjamin Franzke <benjaminfranzke@googlemail.com>

Status

    Proposal

Version

    Version 1, March 1, 2011

Number

    EGL Extension #not assigned

Dependencies

    Requires EGL 1.4 or later.  This extension is written against the
    wording of the EGL 1.4 specification.

    EGL_KHR_base_image is required.

Overview

    This extension provides entry points for binding and unbinding the
    wl_display of a Wayland compositor to an EGLDisplay.  Binding a
    wl_display means that the EGL implementation should provide one or
    more interfaces in the Wayland protocol to allow clients to create
    wl_buffer objects.  On the server side, this extension also
    provides a new target for eglCreateImageKHR, to create an EGLImage
    from a wl_buffer

    Adding an implementation specific wayland interface, allows the
    EGL implementation to define specific wayland requests and events,
    needed for buffer sharing in an EGL wayland platform.

IP Status

    Open-source; freely implementable.

New Procedures and Functions

    EGLBoolean eglBindWaylandDisplayWL(EGLDisplay dpy,
                                       struct wl_display *display);

    EGLBoolean eglUnbindWaylandDisplayWL(EGLDisplay dpy,
                                         struct wl_display *display);

New Tokens

    Accepted as <target> in eglCreateImageKHR

        EGL_WAYLAND_BUFFER_WL                   0x31D5

Additions to the EGL 1.4 Specification:

    To bind a server side wl_display to an EGLDisplay, call

        EGLBoolean eglBindWaylandDisplayWL(EGLDisplay dpy,
                                           struct wl_display *display);

    To unbind a server side wl_display from an EGLDisplay, call
    
        EGLBoolean eglUnbindWaylandDisplayWL(EGLDisplay dpy,
                                             struct wl_display *display);

    eglBindWaylandDisplayWL returns EGL_FALSE when there is already a
    wl_display bound to EGLDisplay otherwise EGL_TRUE.

    eglUnbindWaylandDisplayWL returns EGL_FALSE when there is no
    wl_display bound to the EGLDisplay currently otherwise EGL_TRUE.

    Import a wl_buffer by calling eglCreateImageKHR with
    wl_buffer as EGLClientBuffer, EGL_WAYLAND_BUFFER_WL as the target,
    NULL context and an empty attribute_list.

Issues

Revision History

    Version 1, March 1, 2011
        Initial draft (Benjamin Franzke)
jerry@jerry-Aspire-7739:~$

[SteveRiley]

I am extremely eager for Wayland to arrive. Shedding some ancient X baggage sounds fantastic. The project's progress is rapid and looks promising.

[SecretCode]

Part of me looks forward to Wayland.

Part of me fears (with high certainty) that it will take years for the vast array of useful tools surrounding X to develop equivalents for Wayland - xprop, xwininfo, integration with display drivers, knowledge of virtual desktops, etc etc.

[SteveRiley]

Part of me fears (with high certainty) that it will take years for the vast array of useful tools surrounding X to develop equivalents for Wayland - xprop, xwininfo, integration with display drivers, knowledge of virtual desktops, etc etc.

Perhaps not quite so long as you think? Because Wayland supports running an X server as a client, I don't suspect we'd have to immediately abandon all our favorite tools.

Although it's a year old, Phoronix has an interesting bit about how Wayland will handle virtual desktops and multiple monitors.

[SecretCode]

I am slightly less afraid.

[GreyGeek]

I believe it will be a great improvement.

[jpenguin]

It does sound like it will be better, I too hope the wait until it's stable before they start using it. We won't be able to forward stuff the same as with X though

[SteveRiley]

We won't be able to forward stuff the same as with X though

Most critics of Wayland's lack of network transparency mention this. I'm curious -- is such use of X still all that common? Wouldn't something like VNC be a better choice?

[lo_koraco]

Perhaps not quite so long as you think? Because Wayland supports running an X server as a client, I don't suspect we'd have to immediately abandon all our favorite tools.

All the X utilities are native. There's no porting them to Wayland, as it's (correctly) designed as a dumb display server, nothing else. Stuff will have to be rewritten, or run in a server as client as server to clients compatibility mode. It's less of an overhead than what we have now with X as a middle man and a huge array of extensions trying in vain to keep up with modern demands, compositing and hw accel features being the main focal point. As for your other point about VNC, no tool is as convenient as doing an ssh -X user@server. But, the overhead to using network transparency by running X as a Wayland client is again much less of a problem than running X with all its extensions. The only real problem is Nvidia and AMD, period. I say screw them, but people are adamant about running games in Linux or whatever.

Wayland is a beautiful piece of code.

[Guest]

Most critics of Wayland's lack of network transparency mention this. I'm curious -- is such use of X still all that common? Wouldn't something like VNC be a better choice?

No. I work with X's network transparency on a daily basis and VNC is definitely not an alternative.

[bsniadajewski]

Hopefully Nvidia will get on board with Wayland soon enough.

[jpenguin]

I do occasionally, but not often. The benefits Wayland should bring, far outway this small inconvenience

Most critics of Wayland's lack of network transparency mention this. I'm curious -- is such use of X still all that common? Wouldn't something like VNC be a better choice?