Wayland 1.20 Planned For Release In December

[birdie]

So let me get this straight.. On one hand, Wayland compositors have to reinvent the wheel because of the protocol's minimalism. On the other hand, that's what Wayland want, because they don't want to turn into a "bloated mess" like x11. Can't there be some middle ground in this situation?

No.

Cue oiaohm who will soon come and explain to you that: "Wayland was created for simple embedded devices first and desktop support was an afterthought. Secondly, it's developed mainly by RedHat and they are content with the state of Gnome under it".

[bple2137]

Also, Wayland was released in 2008; it's hardly in it's "early years" anymore.

Meanwhile, bunch of proposed protocols and enhancements that could help filling some missing gaps, are just stuck in code review for months or years (some are never going to make it to the `main` branch). It's either because of complexity of those topics and lack of ideas on how to design perfect solutions (compromises are never allowed), or lack of priority. https://gitlab.freedesktop.org/wayla...merge_requests

It progresses, just very slow.

[MadCatX]

X.org, though it was a fork of XFree86, saw a ton of development in its early years.

X.Org had a lot of slack to pick up after XFree86. More importantly, X.Org is orders of magnitude more complex with its own rendering API, raw hardware access, PCI bus handling(!) and a lot of other garbage that really shouldn't belong into a display system. Wayland is much leaner, hence the there is no need for any hectic development.

[MadCatX]

So let me get this straight.. On one hand, Wayland compositors have to reinvent the wheel because of the protocol's minimalism. On the other hand, that's what Wayland developers want, because they don't want to turn into a "bloated mess" like x11. Can't there be some middle ground in this situation?

The situation is not really as dire as some people like to make it out. A lot of the heavy lifting is handled by external tools the compositors can tap into; such as pipewire for screencasting. There is also wlroots which serves as the de-facto template for Wayland compositors. GNOME and KWin have it worse because they need to reimplement some stuff by themselves. If you remember the early days of hardware accelerated desktop or video playback on X.Org, that has also been an ugly mess for years (and to an extent still is) so it's not like X.Org/X11 were easier to develop for.

[TemplarGR]

Why does it feel like it's already dead Jim? Or everyone is content with the status quo? X.org, though it was a fork of XFree86, saw a ton of development in its early years.

Why does it feel you are an amateurish troll, Birdie? Wayland is a protocol, not a server. Wayland compositors are where the real work is being done.

[JPFSanders]

So let me get this straight.. On one hand, Wayland compositors have to reinvent the wheel because of the protocol's minimalism. On the other hand, that's what Wayland developers want, because they don't want to turn into a "bloated mess" like x11. Can't there be some middle ground in this situation?

Let me give you an example:

TCP/IP doesn't provide SMTP/HTTPS/SFTP/SSH functionality so due to TCP/IP's minimalism. Software like Firefox or Chrome, Postfix, Exim or OpenSSH have to re-invent the wheel to provide all that functionality to TCP/IP.

Do you get it now? Wayland is a protocol that limits itself to the display, not what the user can't or can't do with the display.

Wayland is a display protocol, not a desktop with all the bells and whistles of a full desktop environment.

And yeah, I am too of the opinion that Wayland's approach has been sub optimal, in my use case I need X.org and the programs that run in X.org, and I do not see me in the near future using Wayland due to the shortcomings of missing functionality of DE's under Wayland.

[bug77]

Can it recover from GPU hangs?
Can it update GPU device driver without restart?
Can it support secure attention key login using Ctrl+Alt+Del?

No, but it has resolved the more pressing problem of a branch having a non-inclusive/offensive name

[JPFSanders]

No, but it has resolved the more pressing problem of a branch having a non-inclusive/offensive name

Eggcellent we can all go home and feel good about ourselves.

[birdie]

Why does it feel you are an amateurish troll, Birdie? Wayland is a protocol, not a server. Wayland compositors are where the real work is being done.

Is the Wayland protocol feature complete? I've seen very few people who think so outside RedHat and some staunch Wayland fans here on Phoronix who believe Gnome/Mutter is Wayland. The fact that e.g. XFCE and LXQT don't even have plans to support Wayland due to sheer complexity of doing so doesn't bother them the slightest. The lead IceWM developer outright refused to support it. The fact that KDE is struggling to work properly under Wayland is also quite telling (last time someone on Phoronix tried to run it, they discovered at least 7 criticial bugs).

Lastly it looks like you just cannot stop with insults. This is probably the 20th time you're calling me names. And our brave, honest and impartial moderator, tildearrow will as always neglect your insults and only threaten to ban me. What a wonderful community we have here.

[CTTY]

fixing CVE-2013-2003 as that eight year old CVE for libXcursor never saw its similar integer overflow fix imported into the Wayland Xcursor code

I use sway for two years now. How bad was this vulnerability and why wasnt it merged earlier?
I am a bit afraid I had this whole in my system for 2 years.

Or did it only affect weston?