Announcement

Collapse
No announcement yet.

The X.Org Server Continues Cruising Along As We Approach 2019

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • TemplarGR
    replied
    Originally posted by ThoreauHD View Post
    IRT Wayland, that's like buying a new car with no windshield. I have no problem with new. I have a problem with broke as fk and the idiots that refuse to accept it. I ain't driving a display server that I can't see a single fking thing out of. Nvidia said they aren't doing Wayland. Period. They are 70% of the users/market. Whatever great idea Wayland is, that's exactly where it dies, until you can convince Nvidia to give a shit about that pet project.
    Nvidia is like the 10%-15% of the desktop market, actually... Nvidia is 70% of the dGPU market. The vast majority of desktops use iGPUs from Intel and AMD though...

    So really, who gives a shit what nvidia thinks of wayland? Keep in mind that Nvidia's marketshare will keep shrinking and shrinking as APUs become more prevalent and powerful and AMD's dGPUs regain marketshare...

    Leave a comment:


  • Britoid
    replied
    Originally posted by skerit View Post

    I was kind of amazed how absolute coordinates is something that is not supported. It seems like such a no-brainer. (And it's also one of the reasons why Wine is not supported on Wayland.) If it's things like this that drove Canonical to create Mir, I wish they would have stuck with it.
    Given that Wine still uses Xorg on OSX afaik after how many years, I don't see Wine being ported to Wayland natively. But that's "fine". Wine apps aren't 1st class applications, we don't expect them to be perfect, expect XWayland to stay around for at least 10 years.

    Originally posted by ThoreauHD View Post
    IRT Wayland, that's like buying a new car with no windshield. I have no problem with new. I have a problem with broke as fk and the idiots that refuse to accept it. I ain't driving a display server that I can't see a single fking thing out of. Nvidia said they aren't doing Wayland. Period. They are 70% of the users/market. Whatever great idea Wayland is, that's exactly where it dies, until you can convince Nvidia to give a shit about that pet project.
    But the nVidia driver supports 'Wayland', it just doesn't support Mesa's Buffer system for X, Y, Z reasons.

    Leave a comment:


  • ThoreauHD
    replied
    IRT Wayland, that's like buying a new car with no windshield. I have no problem with new. I have a problem with broke as fk and the idiots that refuse to accept it. I ain't driving a display server that I can't see a single fking thing out of. Nvidia said they aren't doing Wayland. Period. They are 70% of the users/market. Whatever great idea Wayland is, that's exactly where it dies, until you can convince Nvidia to give a shit about that pet project.
    Last edited by ThoreauHD; 28 December 2018, 11:36 AM.

    Leave a comment:


  • aufkrawall
    replied
    It's not true that there wouldn't be any Nvidia driver issues with Xorg, it has severe issues vs. Mesa / FOSS DDX drivers.

    Leave a comment:


  • skerit
    replied
    Originally posted by pininety View Post
    There are some "features" which will never make it into wayland, maybe as a protocol into the compositors but not into wayland itself. (AFAIK absolute coordinates are one of those).
    I was kind of amazed how absolute coordinates is something that is not supported. It seems like such a no-brainer. (And it's also one of the reasons why Wine is not supported on Wayland.) If it's things like this that drove Canonical to create Mir, I wish they would have stuck with it.

    Leave a comment:


  • pininety
    replied
    Originally posted by davidbepo View Post

    thats due to the some other reasons you said and some that you didnt mention like feature parity
    Once's feature is someone elses security bug, unfortunately.
    There are some "features" which will never make it into wayland, maybe as a protocol into the compositors but not into wayland itself. (AFAIK absolute coordinates are one of those).

    Leave a comment:


  • schmidtbag
    replied
    Originally posted by davidbepo View Post
    thats due to the some other reasons you said and some that you didnt mention like feature parity
    That's a compelling argument to favor Xorg, but not compelling to hate Wayland. Also, the lack of feature parity kinda ties in with my 5th point.
    Keep in mind, there are people out there who just simply don't want Wayland to exist; they want it to fail, or at the very least, expect it to.

    Leave a comment:


  • milkylainen
    replied
    Well. The same thing goes for all complex pieces of software.
    Don't get me wrong. I love that Wayland is making progress.
    But it's usually the same ol' story.

    1. "Boo. <Insert large software here> is crap. We can do better!"
    "Our software will be a bazillion times faster and more shiny!"
    2. "Crap. There is a zillion features people have come to expect. This is going to take some serious work..."
    3. 10 years later, still playing catch up.

    I remember when LLVM/Clang started out.
    So many "Boo, GCC is crap, this is sooooo much faster and more shiny!"
    Turns out after all these years of churning to reach the same level of maturity and functionality that:

    1. Performance aspect in different categories is still approximately the same (give or take).
    2. Once all features start adding up, it will become a large piece of obnoxious software.

    It takes time for complex pieces of software to mature.
    X has been brewing (and rotting) for ages. I don't expect Wayland to replace it everywhere just yet.

    Leave a comment:


  • davidbepo
    replied
    Originally posted by schmidtbag View Post
    * Xorg has a lot of sycophants that, for whatever reason, seem to despite Wayland..
    thats due to the some other reasons you said and some that you didnt mention like feature parity

    Leave a comment:


  • schmidtbag
    replied
    Wayland's slow adoption is really just due to a big series of bad luck and oversights:
    * Both GNOME 3 and KDE 4/5 are pretty polarizing DEs, but they're also the best way to get Wayland out there. A lot of people opted for environments not yet (or ever will be) ported to Wayland as a result, which is a pretty significant drop in userbase.
    * Basically any GPUs not made by Intel or AMD weren't compatible for most of this time.
    * Canonical had to waste their time with Mir, which drew a LOT of attention away from Wayland.
    * Xorg has a lot of sycophants that, for whatever reason, seem to despite Wayland.
    * Wayland put way too many necessary or implied features into the hands of toolkit and DE developers, which dramatically slowed down its progress. I don't necessarily disagree with Wayland's decision to be like this, but, it has sort of backfired.
    * Naming the test platform as "Rebecca Black OS" probably didn't do them any favors (though I personally found it to be pretty funny, and sure caught my attention).
    * Most importantly, backward compatibility with programs tied to Xorg has been a major turnoff, due to performance and integration issues.
    Last edited by schmidtbag; 28 December 2018, 02:25 PM.

    Leave a comment:

Working...
X