X.Org is now pretty much an ex-org: Maintainer declares the open-source windowing system largely abandoned Red Hat's Adam Jackson, project owner for the X.Org graphical and windowing system still widely used on Linux, said the project has been abandoned "to the extent that that means using it to actually control the display, and not just keep X apps running." Jackson's post confirms suspicions raised a week ago by Intel engineer …
.....as in doing maintenance on a headless machine over the network....like this:
*
$ ssh -X -C -l your_name_here -p your_port_here remote_IP_address_here
*
I'm still not clear how this is supposed to be accomplished in the new spiffy Wayland-only world. Can anyone help me out?
Apparently 'waypipe' is supposed to be able to do this. I mean yes it works by throwing bitmaps around, but in practice so does most X work now. (Except Emacs, which is probably a major reason to keep XWayland around. XWayland doesn't work with everything, but Emacs doesn't use the modern stuff like systrays etc that XWayland doesn't like, so you should be OK.)
The problem with modern X applications working by throwing bitmaps around is actually a problem with the way that people write X client programs today, rather than a problem with X.
They've forgotten all of the short cuts and efficiencies that were built into X to allow it to work efficiently over bandwidth limited networks, as was the case when 10Mb networks were all you would expect to see.
The original font handling code, for example. It had it's limitations, but by using the font definitions stored on the X server it allowed you to transmit text into remote windows extremely efficiently (a few bytes per character compared to full bitmaps for each character), especially once they built scalable font support into X. By the problem is that you were restricted to the fonts that the server had available locally. The official solution was to create font servers, so that when you wanted to use a font that the X server didn't have, it knew how to find and download it, (although there were licensing issues with making fonts generally available on a network, but that's another story).
Instead, what people ended up doing was to locally compose to a pixmap of the entire window on the client, so only the client had to have the font definitions, and then push the whole pixmap for the window across the network.
This is inefficient (actually X even has some efficiencies that allow you to just send the deltas), and only was acceptable because of the increase in speeds in the network.
Obviously, there are problems with highly graphic intensive applications, such as video handling, but by using client and server on the same system and using shared-memory communication between the client and the server, this could be fixed, but then the 'lets throw the baby out with the bath water' brigade decided that the whole X protocol thing running locally was a waste of resource, and promptly started ripping large parts of the function out for the sake of efficiency, rather than making it work better.
Of course, abstracting the graphics acceleration layer on the local system to allow programs to directly use hardware acceleration is always going to be more efficient, but by using an intermediary interface, such as OpenGL - which the X.org implementation can use, can even make this more efficient.
And yes, there are security issues, but networking security issues have been fixed for other things, and would have been (and some have been) fixed as time passed. The computing world used to be a much more trusting environment than it is now, but people took the time to fix it.
X is an old protocol, but as has been pointed out, does things that the replacement systems simply can't do. I keep trying to explain to my wife that just because I'm not sitting in front of a system that is on in one of the other rooms in the house, it does not mean that I'm not working on that system. Most people just see the down-sides, and never the upsides.
I'm an old hack. I used to be one of the people handling X support for AIX in the UK, and I still use it frequently today, even between Linux systems. And I mean today, as I've been doing remote work on one Linux system and an AIX system at my home from my Linux laptop, and also using VNC to control a Windows system (I don't have RDP on the system, at least I don't think I have), and I can tell you the way X works is much more flexible and efficient, and more pleasant to use than VNC.
Agreed. I'm not particularly impressed by Wayland and its orientation toward local, single-user systems.
I was writing X11 applications at IBM in the late 80s / early 90s: clients, a window manager, graphics libraries (XGKS), and extensions (PEX). I wrote the ddx side for some experimental display hardware. While there were some unfortunate choices in the X11 protocol - specifically, it would have been nice if clients could specify strict or relaxed rendering of wide lines and other primitives to make better use of acceleration - X11 was a rather brilliant piece of work.
VNC is just network framebuffers. It's the sort of remote-display technology an undergrad would come up with. It has its uses, but comparing VNC to X11 is like comparing a pedal car to a Ferrari.
I've never looked at RDP closely, but apparently it's based on the ITU's T.120 family of specifications, and those are just as elegant as you'd expect.
Maybe we spoke. I was in the ADG group of the AIX Systems Support Centre, although if you were using PEX and XGKS, you probably talked to Tim D. if you called in. He tended to grab those calls.
I remember that the accelerated line drawing of the Sabine did not match the strict pixel-by-pixel X11 specification.
I'm still not clear how this is supposed to be accomplished in the new spiffy Wayland-only world.
It's not. Wayland becomes a way of FORCING LINUX USERS INTO A MS WINDOWS MODEL, just like systemd and "other Poettering things"
And, according to the article, X.org is *HARDLY* ABANDONWARE. Last commit a few hours ago, right?
If they need more people to keep X.org alive, I'll volunteer MYSELF.
But... I will *NOT* be *FORCED* into *WAYLAND* and *LOSE* the *ONE* *BEST* *FEATURE* of X11 protocol, the use of the DISPLAY environment variable to run on a REMOTE desktop [or the same one, with a different user context].
that word - "modern". Hey RED HAT, I do NOT think that word means what YOU think it means!
I recently built an application that interfaces a windows clipboard with an X11 clipboard.
Windows side - "push contents to clipboard, pull contents from clipboard".
X11 side - "Send a request to the window that owns the select, wait on a conditional mutex for the message to come in on the X11 message handling thread, read the data length and type from that message, tell the other window to delete it, request the data again, wait on the conditional mutex again, now you may read the data (but don't forget to delete it). Oh, hope that the other program doesn't crash or change selections partway through this process."
So...
This post has been deleted by its author
Windows side - "push contents to clipboard, pull contents from clipboard".
Apart from the fact that Windows probably has to do a fair bit under the hood there's a substantial difference in operating environment between a system with a single, local user and what is, potentially, a networked multi-user system.
"there's a substantial difference in operating environment between a system with a single, local user and what is, potentially, a networked multi-user system."
The phrase of perfect relevance in regards of Linux on the Desktop's problems. Linux is the incredible server OS. But a great server OS does not make a great desktop OS, no matter how much supporters wish to push it that way. The usage paradigms, and expectations, are fundamentally different but too many supporters completely believe otherwise.
Unix on the desktop never went anywhere. Xenix on the desktop too. The only reason BSD went anywhere - MacOS - is that a company, Apple, adopted the code as a base but not as a usage paradigm, they wrote an entirely new, custom UI and designed a system for average desktop users where multiple login credentials is a bonus, not the highlight. The system was designed for user experience, not for the strict compliance of making the hardware virtualized between all those independent login users with total isolation; one user on the system at a time, full access to said system during use, with administrator privileges looking over your shoulder in case something funny looks like it may happen.
Linux's main advantage, multiuser, is a hindrance in every day, average-user desktop use. We don't try to force the dozens of variations of Unix on mum's computer because we know that's it's compete overkill and inappropriate, but we're willing to push Linux on her because it's such a great OS that, never mind the inherent complexity, it'll work great!
So now X is being abandoned. One of the reasons? Too complex to do the job at hand - create a great user experience in a structure that is also easy to design with and maintain. But nobody will want to discuss the elephant in the room, and I'll be downvoted for sure :D
I'm sure that Windows does do stuff under the hood, mainly with converting to different formats instead of requiring you to provide each format yourself. Good, that improves interoperability with others applications and reduces headaches for every single developer who ever touches the system.
And as to being multi-user, if X11 can't handle giving each user session it's own instance of the clipboard system, then what horrible things is IT doing underneath?
I'm sure that Windows does do stuff under the hood, mainly with converting to different formats instead of requiring you to provide each format yourself. Good, that improves interoperability with others applications and reduces headaches for every single developer who ever touches the system.
Let's not be too kind to windows here, if you're talking Win32 you still have to do all that shit....
And as to being multi-user, if X11 can't handle giving each user session it's own instance of the clipboard system, then what horrible things is IT doing underneath?
Exactly! Lets be fair too, Windows hasn't really been 'single user' since ME, the NT line is quite happily multi user, even if the functionality is restricted in the desktop versions. If you lock your machine and another user logs on they ain't pasting your clipboard contents. :)
X11 is just shit. It should've been taken out and shot years ago. There's a reason Apple went with display PDF for their desktop UNIX (Okay, that reason is NeXT had long ago looked at X and gone eugh ;), but still....)
X11 side - (for the clipboard)
Not that hard, and a bit more flexible than it is in the windows world. In theory, within the X11 world, you can store multiple data types on the clipboard (each with its own owner), and not "just one thing" (they don't even have to be for the same object or application, etc.). Additionally it's really not all that hard to make use of in the X11 world. Many clipboard APIs and libraries exist for just about every toolkit I'm aware of. It's more or less a "solved problem" with a lot of sample code available.
I use a background thread and some sync objects (in my own toolkit) to make it appear to work synchronously from the main thread. Copy/Pasta performance is pretty good. The UI is basically seamless, no stuttering or weirdness. code is simple (synchronous call similar to Win32 API in that regard) with the toolkit doing the async stuff in its own thread as needed.
There is a nice command line example 'xclip' that's not all that hard to follow, which lets you access the clipboard from the command line. So if you need clip functionality and are NOT using a toolkit (or are writing your own), this is likely to be very useful to you (it was for me, a long time ago):
https://github.com/astrand/xclip
Question: why abandon something that works and is in long term maintenance mode?
2nd Question: who here has NOT read Arthur C. Clarke's "Superiority" ?
X11 has a LOT of legacy stuff in it, which I expect is NOT used all that often (except x sample applications like xclock etc.). And better docs for the more recent extensions would be nice, like glx, shared memory extensions, and so on...
Freetype and related stuff are actually add-on libraries.
THAT being said, it is NO excuse to go and ABANDON a working system for "something that more resembles the way Micros~1 does it" like WAYLAND, *ESPECIALLY* when you *LOSE* *CAPABILITIES* that are EXTREMELY IMPORTANT!!! You know, like the 'DISPLAY' environment variable and REMOTE X server support.
/me uses pluma on an RPi to edit files NEARLY on a daily basis for a major client of mine, using 'DISPLAY' to edit source files on a Linux workstation across the network, often WIRELESSLY. It's a touch screen system, like so many others out there i bet. NO WAY could I do editing on that touch screen, either, and it needs to have the application RUNNING ON THE TOUCH SCREEN to test the code I'm working on. I can't imagine the HORRIBLE INCONVENIENCE AND SLOWDOWN that a LACK of remote X11 capability would cause...
(yeah NOBODY's doing that remote X server thing... right? What, HALF the people who read El Reg do that at least OCCASIONALLY? That sounds about right)
Does Wayland provide network transparency yet, or are we still back in the world of Suntools (or possibly faking it by some equivalent of VNC)?
Oh, I'm sure the Wayland folks will tell you they're SERIOUSLY thinking about starting to add that functionality in the next 5 or 10 years, if they aren't busy with something else, and besides, just WHY are you doing something foolish like that anyway, and they can't figure out how to run X11 as a SystemD service...
My impression has been that the Wayland development culture is quite juvenile in character. They have rejected contributions which provide missing features, as for instance equivalent xrandr functionality, and they seem more interested in exactly who does and does not get to share the secret decoder rings or have access to their private tree fort. After over 12 years of development, and *still* not a replacement for X, expect another maybe 10 years for the missing functionality to become available.
"expect another maybe 10 years for the missing functionality to become available."
Yes, Wayland started gaining traction 10 years ago according to the article. 10 WHOLE FRIKKEN years. And it's still not ready for primetime! It's still not ready to replace X, for all of the X disadvantages. And I too am one those who frequently runs X programs remotely. Remoting into Windows PCs always feels so clunky because the ONLY option is get a full remote desktop, not just the program I need to run.
I would just leave the code lying around somewhere. Someone is bound to pick it up in the future and restart development. THAT'S the power of open-source: anyone can take the code and start adding stuff or refactor it. If the project's dead you can usually apply to become the maintainer.
Wishful thinking... someone will just rewrite it Javascript before thinking that half Rust, half Go looks nice, before replacing part of it with Kotling, and adding modules in TypeScript. With Python to glue it all together.
Maybe it's just time to let X die and replace it with something better, designed for the XXI century....
"Maybe it's just time to let X die and replace it with something better, designed for the XXI century...."
Despite all the downvotes, that's not really such a bad idea. After all, Linus took Unix and wrote a work-alike based on the system behaviour. What's so bad at starting from the ground up with the X specs and starting from scratch with a modern work-alike? Get rid of the ancient cruft and add new features, optimising as you go.
Get rid of the ancient cruft and add new features, optimising as you go.
I'm not sure there's a LOT of "ancient cruft" there. Some of that still has use, especially for running a GUI application on an embedded system (let's say something in the 100's of Mhz clock frequency ballpark) across a network (let's say a WiFi G network with lots of interference) and still get reasonable UI performance in which the application is still "usable". not like you're playing videos or anything...
I like to think of that "ancient cruft" as "well tested bug-free code".
> I like to think of that "ancient cruft" as "well tested bug-free code".
Totally agree.
Besides security issues that come along and need patching, code does not degrade with age and so "ancient cruft" is a newbie coders way of saying "I dont want to learn how it works because its hard getting up to speed so I will just write my own version that breaks all the cool stuff and implements only 50% of the original features because thats all thats needed to run Steam, then find ways to force everyone to use that while convincing everyone that its progressive.
There is a case for chucking everything out and starting from scratch, feature for feature.
There is a case for deciding if a bit of functionality SHOULD be in this code or outside. This is part of the Unix design philosophy where you can strip out features into their own programs thus making everything simpler and more modular. Does wayland have this? A Wayland Networking daemon that gets started on demand when networking comes into play? It could allow anything from single windows to whole desktop networking (basically a whole X session over the network) and can leverage another program to set up encrypted tunnels. Nope, that doesnt run steam so nope. All it needs to be is a proxy, forwarding Wayland events to and from each systems compositors. The compositor need not care that its over a network, just serve the standard wayland protocol.
"Ancient cruft" that may actually be there is most likely hacks put in to work around an issue with hardware that nobody uses anymore, even Linux has that.
"I dont want to learn how it works because its hard getting up to speed so I will just write my own version that breaks all the cool stuff and implements only 50% of the original features because thats all thats needed to run Steam, then find ways to force everyone to use that while convincing everyone that its progressive."
Good grief. have you been talking to Poettering?
“ I would just leave the code lying around somewhere. Someone is bound to pick it up in the future and restart development.”
I think MSFT perhaps already did. It’s how they provide enhanced mode access to Hyper-V hosted Ubuntu (complete pain in the arse to get working with 20.04, btw, but rather nice once working).
> When a response to Jackson's post complained about issues with Wayland – such as lack of stability, poor compatibility with Nvidia hardware, lack of extension APIs – the maintainer said that keeping X server going was part of the problem.
IME lack of stability has been more closely related to design flaws rather than amount of development effort.
Xwayland stacked on Wayland really just feels like a re-implementation of Xorg which doesn't really phase me. Either way I am reliant on X11 for my WM and many tools I use and products I develop.
Hopefully the "new stuff" it is more maintainable for the current generation of developers and we can just carry on using X11 like nothing has changed ;)
Xweston is basically a single fullscreen Xwayland window so you can ignore the rest of the "difference"
Perhaps one day we will look into replacing X11 with something completely different. Especially since as it is now, RDP is a little better across the network which is fairly annoying because the network feels more like the UNIX-like domain. But for now, X11 reigns.
Either way, Xwayland will likely be a nicer experience to Xsun.
"Xwayland stacked on Wayland really just feels like a re-implementation of Xorg which doesn't really phase me." "Xweston is basically a single fullscreen Xwayland window so you can ignore the rest of the "difference""
So I just posted how I hope X.Org keeps on being at least considered non-abandoned, until Wayland finished getting better. I guess if this is available, as long as the performance is acceptable that'll work for me; if wayland keeps refusing to add some functionality, I can get by with Xwayland or Xweston, and the functionality will sit in that layer (may not help with xrandr, but I'm not using that personally.)
"Especially since as it is now, RDP is a little better across the network"
I'm going to guess that you've not encountered Xrdp yet... https://xrdp.org
It's a pretty decent implementation of RDP to reach an X11 based server. There's a bit of a weird default where it uses RDP over the network and then VNC over loopback but there's also a drop in Xrdp Server which seems to provide native RDP access
I've been using it for a couple of years on Ubuntu 18.04 and it's way nicer to use than VNC. The shared clipboard works well and it also supports some peripheral sharing and drive mapping, neither of which I've tried.
Up until lockdown I was able to access my work Desktop physically in the office and then pick up the same session at home. Since march I've been using it 100% remotely and it's been fine.
There's one little nuisance which is that sound support had to be added via a 3rd party pulse audio plugin. It's well documented though.
The default config shipped with 18.04 doesn't actually work, which is a shame but it's not too hard to poke it into shape.
XWayland stacked on Wayland
How to actually make something like this WORK... (but will the alleged CHILDREN working on the Wayland project ever consider such a reasonable proposal from an OLD FART like me?)
* XServer uses X.org implementation "as-is" running like a subsystem layer [my read-up on XWayland just now suggests it is not really like this at all]
* XServer runs independently, NOT invoked by Wayland (so you can have it listen on a TCP socket, just like always, NOT like XWayland currently is, if I read it correctly)
* X11 Video driver and glx and shared memory extensions do the actual work, implemented by X.org as a video driver, and not necessarily a part of Wayland [but with X11 hooks in Wayland as needed]
* OpenGL lib similar to what NVidia does, that does more direct video things at a low level, so you get equal [and possibly better] performance with X11.
* XVideo hooks via the X11 video driver.
This would essentially be 'wayland as a soft video driver'. It would support remote X11, and existing X11 applications without modification. It would essentially use the SAME X.org server code.
But here's what it would NOT do: It would NOT force EVERYONE ELSE to CHANGE, thereby giving up tried/true for the NEW SHINY, sorta like as described in [mentioned again] Arthur C. Clarke's "Superiority".
> RDP is a little better across the network
How? Last time I used RDP I had a full desktop with notifications and scrollbars because it couldnt fit on my monitor. All I needed to do was add a machine to DNS, RSAT tools was broken on my laptop for some reason and I probably could have used powershell. It would have been nice to simply launch the DNS dialog on the Domain Controller and have it work as a normal window on my machine. But no, I had to have a HUGE RDP window and wait for RDP to log me in.
I sure hope X.Org doesn't die off yet! As thx1111 alludes to above, Wayland is yet another X.Org replacement that's got about 90% of the functionality... that old saying "It takes 90% of the time to get 90% done, and another 90% to get the other 10%?" Well, Wayland's been "almost" ready for like 5-10 years, but they just say "You don't need that other 10% functionality" rather than even trying to add it.
For example... I run a remote app or two daily. Wayland doesn't support running an app over the network. OK, so I should use a remote desktop? a) I just want one app... b) Due to the discussed better security, it doesn't support running a remote desktop either! (That would involve being able to read the whole screen, which Wayland disallows for security reasons, and effectively inserting keystrokes and mouse movement, also disallowed for security reasons.) God forbid they give you an option to run an app as a privileged user or something to be able to do those things.
No equivalent to xrandr, because they decided you don't need it. (If people want it, why not add support? Maybe not literally xrandr command, since that "x" is for "X", but some way to adjust the screens!)
As already discussed, it still has poor support for certain video cards -- which you'd think would mature, except Waylands already been out for 5-10 years. It's premature to eject X.Org for this reason alone. It's also apparently unstable for some.
X.Org is quite mature (both in the sense of "old" and in the sense of having many many bugs worked out of it over the years) so I don't expect necessarily extremely active development since they're not adding major new subsystems to it right now. But it's apparently still getting patches and updates; hopefully, if redhat decides X.Org is dead, someone else will take over at least hosting it and collecting/vetting the patches.
In your discussion of Wayland a few things struck me. You say that features in X are not implemented as the Wayland devs have decided that they are not needed and or undesirable.
Does this attitude not strike a chord?
It seems to me that Gnome devs have a similar attitude of "We know what's best, take it or leave it". You know, like hacking away at Nautilus until only a shell of functionality remained.
It could be arrogance, immaturity or even that they don't give a damn and are writing this stuff for themselves and no one else. Whatever it is they are not making any friends.
Indeed, this does stink of GNOME attitudes.
I think there's a fairly hefty number pair developers from both projects who are from RedHat or Canonical - it would be interesting to see a breakdown of just how influencial these firms are in steering these changes.
The GNOME project is definitely heavily invested in Wayland now and their gung-ho approach to UI design is already causing a major knock-on problem in other GTK based desktops and applications (I won't repeat my massive rant now). A problem I'm now dealing with in MATE desktop to do with hamburger menus embedded in pseudo-titlebars which are actually drawn by the application and don't integrate with the desktop theme all stems from work that GNOME began in pursuit of support for Wayland!
" "I'm of the opinion that keeping xfree86 alive as a viable alternative since Wayland started getting real traction in 2010ish is part of the reason those are still issues, time and effort that could have gone into Wayland has been diverted into xfree86," he said."
As others have said above, his statement here is false. I think Poettering simply doesn't like X11.
Exactly.
There may be gradually growing developer effort going into Wayland, but, unless I'm much mistaken, there are still rather few "production" (rather than "testing" / intermediate release / short term support) distros using it rather than X as yet, several years down the line.
X has served us all very well.
One of the good things about the open source community is that anyone is free to come up with an innovative way of doing things, and, if it works well and appeals to developers, it will blossom. But, at the same time, it is also an evolutionary race, and something new will only occupy a niche if it is deemed useful and usable, and sometimes alternative solutions can happily co-exist.
So, maybe, at some point in the future, Wayland will gain a larger share, but not just yet.
I reckon X will still be around for a good few years more, mind you. Whether we are still saying that in 2030 might be a different question, however.
Complains about X11 being over-complex and difficult to work with go back to the 90s. Even at that time, there was a general feeling that X11 should be torn down and built again. Many wished that NeWS had been open-sourced so that it could have been the dominant UNIX-like windowing system. Hence, I think we are well-passed the point where the problem with X11 is "not enough people working on it".
There were many reasons for X11's success and longevity but IMHO a large part of it can be laid at Bob Scheifler's door. Back in the day, I was very, very impressed by how he found the time to be on the user's mailing list giving help to complete tyro's like me, whilst also running the development project, coding and writing documentation. His tolerance and wisdom were much appreciated.
(Can be contrasted against other important developers whose public approachability leaves something to be desired!)
> But the people there back in the day didn't take actual offence, it just seemed like it.
Bollocks to that. I remember an engineer being fired from Boeing for being an arse in a certain Usenet group. He went trolling for a few days, most of us killfiled him but someone decided to report him to Boeing security (he was posting from their network), who promptly investigated and came back with an apology. Another Boeing engineer posted a few days later with details about the guy's fate.
There's always that outlier, but people who act like that are far more common today. In the Usenet I remember (starting in '89) there were flamewars, sure, but people were usually arguing over facts or positions only people around them would care about (emacs vs vi etc.)
There weren't crazy conspiracy theories or people who subscribe to facts from a completely alternative reality from the one most of us inhabit. I remember the green card lottery thing spammed across all the groups and realizing that was the future once people outside the academic and high tech/defense worlds that occupied the internet at the time were outnumbered 1000 to 1 by ordinary people.
I remember the green card lottery thing spammed across all the groups and realizing that was the future once people outside the academic and high tech/defense worlds that occupied the internet at the time were outnumbered 1000 to 1 by ordinary people.
Ah who could forget Canter and Siegel ...the inventors of spam. Mid 90s (ish). Those were simpler times, but yes definitely sign of the things to come.
As a non-techie/coder type...I've used Linux since 1996 on ALL my pcs in the home. I now live 9mnths of the year in France but still have a house (in which our son lives) in UK. I have a self built VPN between the two. About once per month I invariably find I need to run graphical app in UK on my French desktop 750 miles away.
Tried to do this using Wayland but it seems I need to transmit a whole desktop! As my French link is rural broadband @13mps this just isn't an option. Using X over SSH works a treat. Evwry time.
Until Wayland can do all X is capable of, we need to mobilise and support X
"@13mps? Oh the luxury! In my corner of rural France the stagecoach service is faster than our "broadband".
Yep, I was thinking this as well. 13 Mbps no longer qualifies as "rural" in France but more as center of major city broadband !
Posted from my very rural 2.5Mbps broadband :)
I get really annoyed by the attitude that if it's not under active development it's not worth supporting.
Take the Thunderbird mail client, for example - it's got all the features you need in an email client. No I don't want to edit HTML, or interface with the coffee machine, or... I JUST WANT TO READ MY EMAIL!!
Same with X - *my* use case is that I have a number of Linux boxes, and it's sooooo convenient to open xterms on other boxes and work on my on machine with big dual-headed monitors. Contrast that with the work I have to do on Windows boxes, where I have to RDP in, open a whole damn desktop. Or find out that the number of remote logins has exceeded the license limit, then ring up to get other rude users who stay logged in when they're not actually working to log off. And stop fixing my security - I know not to use anything vulnerable when the network is open to the Internet. But on my own secure network X works just fine. So please stop making it so hard to disable the "nolisten tcp" in X.
And what about grub versus grub2 - grub had a config. file you could read, and I could remember the boot commands from memory. Now there's a huge unreadable grub.conf file - computers are good at pre-processing text. I'm not. And what do you get in return? Dinky progress bars and graphics - in a boot manager FFS!!!
</rant>
That feels better - now for a nice hot coffee :)
I'm in agreement with your on "it's not dead just because nobody's making new releases". I can't see how new features would need to be added. Perhaps simply making sure it compiles with new library releases might all most people would want or need.
> So please stop making it so hard to disable the "nolisten tcp" in X.
It's been years since I needed to disable "nolisten" with X11 forwarding via "ssh -X" (or "-Y").
Agreed grub2 is horrible. After an update my grub screen suspiciously changed font and shade of color, unsettling because I had to enter my crypto password.
After that I switched to systemd-boot, with an ultra simple config of only a few lines, new menus, CLI entry. There's a little bit of manual involved in kernel updates now (~1 minute), but I prefer that anyway.
"And what about grub versus grub2 - grub had a config. file you could read, and I could remember the boot commands from memory. "
I use a custom version 40_custom in grub.d which can have (more or less) the same complexity and the readability of the original GRUB configuration. You need to delete some of the other files in grub.d: 10_linux, 20_linux_xen, 30_os-prober, and 30_uefi-firmware (I save them in a backup directory in grub.d first, just in case) that may override 40_custom or add to grub.cfg when update-grub is automatically run after kernel updates. The biggest downside is that you have to edit 40_custom to match the kernel version when an update changes it and then run update-grub before you reboot. And you have to manually add any OSs for dual boot just like the original grug.cfg. I haven't done this with UEFI boot with GPT partitions yet so I have no idea at this time what if anything else is required to make that work.
I can post an example of a custom multi-boot 40_custom if you like.
Take the Thunderbird mail client, for example - it's got all the features you need in an email client. No I don't want to edit HTML, or interface with the coffee machine, or... I JUST WANT TO READ MY EMAIL!!
It would, though, be kind of nice if Thunderbird could have an actually working implementation of RFC3676 format=flowed ... you know, one that correctly distinguished between wrapped and uwrapped lines in both new message text and quotation text. Thunderbird did rather champion format=flowed when it was new and it would be nice to think that they, if anyone, got it right.
That's for text/plain Content-Type, of course, and the kids these days do mail in HTML, for some reason, so it seems that nobody much cares about RFC3676 any more. I don't want to see HTML formatting, I JUST WANT TO READ MY EMAIL!
I couldn't agree more. What cripples Linux on the Desktop is that we develop some of the best software seen on any desktop, with exacting care given to human factors, how to minimize the number of clicks or keystrokes required to accomplish any task that is the product of the combined talent of some of the best in open-source. And then the developers move one. The project is left to be contributed to, or to have the next major release put out by younger developers with great desire to get the latest gotta-have widgets in whatever it is -- but that have lack understanding of how everything fits together -- but the new widget will look cool... And the great new next best thing to sliced bread is pushed on the Linux community -- only to find it half-baked compared to what it replaces.
That has been the death-nell of many open-source projects. Anyone remember to promise of KDE4 in 2008 when it was pushed out as the default desktop on SuSE 11.0 (a decade of feverish fixed that never really got there and was scrapped for Plasma)? Was there anything broken with KDE3? No, it was one of (and still is) one of the most well thought out and function desktops every made.
Or Gnome3 (where Gtk+3 style elements changed monthly for a decade with each release before being scrapped for Gtk+4)? Was Gnome2 broken? No again ... and we have lived for a decade with applications trying to port from Gtk+2 to Gtk+3 only to find toolbars that spill-off the side of the window because someone thought it was a good idea to remove icons size and spacing as a programming element and turn it over to .css based themes... (See supra - style elements changing monthly further exacerbating the situation)
Or GRUB/GRUB2 mentioned in the comment this is a reply to, or the myriad of browsers and other apps that lurched from one backend to another.
And now we have X and Wayland. The first thing anyone does when they want to push the community to the latest gotta have (whatever) is claim the old one is "broken". Yes, there is politics in open-source development. Set that aside for a moment. The one thing that was correct in the article is that X is a problem for Wayland (and vice-versa). Not that they conflict, but that they both require the same developer talent to create. So long as there is X, Wayland will miss out on talent. And since there is Wayland, X will be starved of talent, releases will be less frequent, etc.... And the end result is Linux desktop is again littered with uncertainty on what business could expect a year from now, again.
There is nothing wrong with X. Wayland isn't all there yet, but with articles like this there is a push to move to it, readiness be damned. (just over look the delay or lack of full implementation for Nvida, or the annoying second screen refresh when accessing a virtualized install) That is the Alkalies-heel for the Linux desktop. Will it get there? Sure, after using a Linux desktop since Mandrake and SuSE 7.0, there is no doubt in my mind it will get there. The crux of the adage, it "if ain't broke -- don't fix it" problem for the Linux desktop is that is probably 300-500 bug reports and commits and 2-3 years away. And that is something you cannot ask the average desktop user to do. The end user is just left wondering why what worked six months ago doesn't anymore and no idea how long it will take to get fixed.
"Some distros still use X.Org by default, including Ubuntu. Wayland is also included, but users have to specifically select it."
I believe that was true for 18.04, but for a fresh 20.04 install so-called "Wayland" (actually XWayland) is the default (*)
XWayland is a Wayland compositor (an implementation of the Wayland protocol) that also has an a modified X server module
so that X apps can run. However, the "security" of Wayland (vs XOrg) is lost when using X apps as follows:
"A Wayland compositor usually spawns only one Xwayland instance. This is because many X11 applications assume they can communicate with other X11 applications through the X server, and this requires a shared X server instance. This also means that Xwayland does not protect nor isolate X11 clients from each other, unless the Wayland compositor specifically chooses to break the X11 client intercommunications by spawning application specific Xwayland instances. X11 clients are naturally isolated from Wayland clients. " [ https://wayland.freedesktop.org/docs/html/ch05.html ]
(*) Apparently, in some hardware setups XWayland performs terribly for games. Ubuntu detects at least some of these cases and then defaults to XOrg.
https://askubuntu.com/a/1258280/723997
I agree with Grumpy Rob. I develop on a Linux box running XDMCP and VcXsrv on windows (actually I have several *nix boxes and XDMCP lets me select which one to log in to). Don't believe it when people say that RDP or XRDP or VNC is "just as good or better". Performance-wise remote X is about as near to native speed as you're going to get, and it's stable too.
"issues with Wayland – such as lack of stability, poor compatibility with Nvidia hardware,"
Well, that's an understatement. I just tried Wayland on Kubuntu 20.04 with an nVidia GTX 650Ti and I get as far as the splash screen, after which the machine just locks up and needs to be power cycled. And the reason I wanted to try Wayland, is because nVidia and KDE/X also sucks:
https://forums.linuxmint.com/viewtopic.php?t=180972
https://askubuntu.com/questions/408424/kde-sometimes-glitches-on-wakeup
Bug goes back at least six(!) years.
Now I love Linux deeply and I've been using it for well over twenty years, but this is one of the reasons the Year of Linux on the Desktop will likely not happen this century. And with Linus Torvalds I say: "F*ck you, nVidia".
I'm wondering why people are complaining at X11 programming. Sure it was really messy, but no-one would access X11 directly, back in the early 90s.
Everyone would go through upper libraries, the name of which I forgot, that made windows programming A LOT simpler and straight forward ... BTW, where have those libs gone ?
Everyone would go through upper libraries, the name of which I forgot, that made windows programming A LOT simpler and straight forward ... BTW, where have those libs gone ?
Few odd times I had any need to actually write anything for X11 myself, it was more or less go via Tk or hook into Motif. Libs are generally available for pretty much most languages and scripts.
"Few odd times I had any need to actually write anything for X11 myself, it was more or less go via Tk or hook into Motif."
Aahh, Motif, that is the one I last programmed on in C. TCL/TK as well of course, which made it, unlike X11, a walk in the park to program with ...
You might be fed up of X11 being old and crusty, but it's all we've got until Wayland is feature-complete. Which it is blatantly not.
Without a working and reliably-deployed UI, the desktop will be owned by Microsoft. Or maybe the point is to ensure that the only way to get a reliable Linux desktop is to buy a RedHat desktop license?
I dont mind using Wayland, well its just a protocol what I really mean is I dont mind using a wayland compositor as long as:
- That compositor never makes itself a requirement. Any application should be able to connect to it.
- It supports networking like X did but maybe in cooler ways.
The one thing I will hold on to X.Org for is to avoid stepping back into 2002 and having to use VNC just because I want to run a graphical program on a headless server or a VM.
The elegance of X is amazing!
The concepts of having an application leverage one or more X Servers (serving resources like Keyboard, Video Display, and Mouse) is wonderful, facilitating a simple multithreaded binary for multiuser applications & games is fabulous! The smooth integration into alternate operating systems of remotely running X applications has always been a bonus.
Sure, it was slower than SunView on a workstation, but it is certainly faster than windows apps with properly written native X applications (without the cruft of overweight took kits.)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
