buillding error xorg-server 21.1.4
I get the following error trying to build version 21.1.4 same build dependencies used for 21.1.3 (4/13/24) but different versions gcc 14.2.1+r134 glibc 2.40+r16 binutils 2.43+r4 libtool 2.5.3 arch linux like system without systemd ../xorg-server/os/access.c: In function ‘siHostnameAddrMatch’: ../xorg-server/os/access.c:1873:27: error: assignment to ‘const char **’ from incompatible pointer type ‘char **’ [-Wincompatible-pointer-types] 1873 | for (addrlist = hp->h_addr_list; *addrlist; addrlist++) | ^ https://termbin.com/fagc (more output before after error)
Re: Xrandr and external monitor
On Friday, January 3rd, 2025 at 12:49 AM, H wrote:
> Forgot to attach the file but doing it now.
I have a shortcut function in my zsh which looks for (EE) in Xorg.log
The first thing I would do after seeing this log is to install the xorg xf86
packages for fbdev and vesa then attempt to run again and rescan for errors.
Subsequent errors and warnings would be meaningless after the first. In the
arch ecosystem those pkgs are marked as xf86-video-{fbdev,vesa} on debian/rh
may have slightly different names.
Re: Xrandr and external monitor
> I just discovered running xrandr --listmonitors that, while the internal > laptop monitor is labeled eDP-1, xrandr identifies the external monitor as > using DP-1 although it uses the laptop HDMI-connector. Further, while mate > hardware display monitor also identifies it as using DP-1 - possibly based on > xrandr output - the monitor itself tells me it is connected using its HDMI1 > connector. xrandr does report the correct display resolution, though. Are both ends of the cable hdmi, no adaptors? Usually hdmi - hdmi or dp - dp seem pretty accurate, while when you go from one to the other they are a bit unreliable. Is the monitor on before you boot, or are you connecting while the system has booted? Pretty much all guis use xrandr output. Since you are not listing the make and model of the monitor do a search and see if firmware are available specific to this monitor. Also which kernel are you using (different versions tend to behave a little different). In most cases the chipset in the monitor provides the firmware needed to properly identify itself in the system, but I've seen many deviations. Do you also have svga plugs on both? Last question, are gfx native intel/amd or is there something like nvidia involved? It is not a good day to be mean to a stranger but tomorrow I may say "ask nvidia then " :)
Re: looking for help on multimonitor setup
> have you run xrandr to see what modes the external screen is offering - which > is claims is native etc.? > - Codito, ergo sum - "I code, therefore I am" -- > Carsten Haitzler - ras...@rasterman.com This is pretty much the answer I would have given as well, if xrandr doesn't show your expected resolution in the list then either the monitor is not recognizable or not sending adequate firmware/identification. Possible glitches can be connectivity, vga, hdmi, dp, dvi-d ..etc. Also, is the monitor plugged before boot or after? It does make a difference with some connectictivity combinations. Same monitors connected via different port behave/identify differently. For those who get frustrated in the terminal formatting a proper instruction there is the arandr gui that works anywhere X is running. It is like the gui side of xrandr, it will even work when starting a terminal without a wm. You visualize how you want the setup arranged and save it into a script. Then look at the script it saved, that is the proper xrandr command that you can add to some autostart script on your wm/dt Most of the headaches I've discovered are cables that are also adapters of one connection to another, as well as tv (spec) used as monitors faking actual resolution and/or size. I don't think there are DTs that can't comply with this because even the oldest most abandoned WMs I'v tried all work with xinerama support. But you are confusing using multiple monitors issue with the issue of properly identifying it. Similarly on wayland there is wlr-randr but in my experience what you get on X you get on wlr
Re: Question about the future of Xorg
Michel Dänzer wrote: > On 2025-06-11 03:53, Felipe Contreras wrote: > > It does seem Enrico Weigelt was the only one, so why wouldn't I > > contribute to XLibre instead? 1 developer is infinitely more percent > > than 0. > Enrico clearly demonstrated in the xserver project he lacks the technical > competence to lead > Xorg in any useful direction and the social skills to participate in a > community project. Don't > let that stop you though. Although it is not polite to speak of people who are banned from the discussion, it appears to me a small group of people just want to lead and own the "next thing" as intolerant dictators without the funds of IBM to pay people to allow it to be dictatorial (as all corporations are, nothing special here). I can fork 21.1.18 as well, instead of old news, and own it, dictate who participates and who doesn't, but the best I can expect is to still have 21.1.18 still functioning 5 years later. Too much noise without much purpose is my assessment
Re: list banning
I am sorry to say that this would have been better not answered than answered the way it did. This is as if some bully comes up at work and spills hot chocolate on your pants, you run to the men's room to clean it off, he follows and steals your pants. Then you get fired for running around the office half crazed screaming non-sense on your boxers. Mission accomplished. Dear Alan, with all respect for your work and contribution, had he used Bcc or a single email would he have not been banned and have all his work from git removed, therefore the merge request you approved and merged not have been reverted? Maybe you should refrain from answering as well, as I said before and someone liked, what happens in IBM stays in IBM. Most of us are mature enough to see all that we needed to see.
Re: Question about the future of Xorg
Robert Funnell wrote: > I would guess that you've just thrown away whatever chance you had of > getting cooperation from the developers. (Is 'citizen journalist' a way > of saying 'unaccredited journalist'?) Unless you have been in a freezer since the mid-80s someone with a blog or website publishing "articles" is a journalist. A journalist is not only someone who some multinational media corporation has branded their employee as such. And at times you hear or read those "accredited" journalists and you ask yourself, why is someone paying them to do what they do, they can barely tie their shoe-laces. Why are we still having such a hard time with independent and unguided free expression and sharing of ideas and opinions without corporate guidelines and agendas? > - Robert On the other hand trying to get statements from people whose weekly income depends on the lack of free expression (as representatives of their employer) ranges from insensitivity to being outright rude and obnoxious. It is rather obvious what is going on here, you don't have to twist peoples' arms to get them to state what you want stated, when it is not officially announced. That's not journalism
Re: Question about the future of Xorg
I don't know whether it is relevant or not and whether it helps anyone understand the practical aspects of it all. If you start a wayland session, wlroots, wm, .. inside an X environment, and run xrandr/arandr in it through Xwayland it shows as the monitors available being X11 (that's how Xrandr identifies the screen it is called to modify). If on another tty you start a true wayland session and within it run xrandr/arandr it identifies the true monitors (hdmi1 dp2 .. whatever) but if you attempt to alter resolution/size it fails, even though an X session is running inside the same system. wlr-randr or something similar can and affects the true wayland session. Whether it is the protocol, how the environment is set, whether it is a matter of authority to modify output .. One thing that makes me scratch my head is how for years distro people were saying that is not secure for a user to own/start an X server, it should be a root or other higher authority starting and then the user can run things within it. But if you made the conscious decision to oversee those security "issues" and were able to start as user, you felt a bit uneasy as doing something wrong. With wayland, if a seat daemon provides a seat to the user, the user can start several sessions. Another matter, I see avoidance of mentioning the word wlroots .. is there a functional alternative to it? I haven't found one.
Re: Question about the future of Xorg
On Tuesday, 250610, Olivier Galibert wrote: > The problems are the reasons why Wayland is inferior, e.g. the integration of > the window manager into the server (making customization so much more > difficult) and In the past while discussing functionality with core devs of a wl wm their refusal to tackle an alternating behavior of some usually X very core choice in wm, was misunderstood. But what you state here explains perfectly well why a wm designer will not go to war with wlroot (example) choices to alter behavior. > the balkanization that results from it Hey, watch it! It is hard to digest why so many people here and at this level will mention "desktop" when what they are talking about is the layer below them, the wm. It brings memories of the famous reaction by Poettering "what do you have against disabled people" back to the foreground. My main negativity source with wlr-stuff is that I can't as easily shift from user to user to user within one session and run the same or other applications as different user. To just run the application containerized as if it is a different session prohibits communication of (clip cut and paste functionality). Say you have your work split as projects and each project has its own user, so "sudo -u proj34 abiword" and "sudo -u proj25 mercury" can't share their clipboard, test or binary. There are ways to cheat this and have a common access saved buffer but it is too much of a hack and hastle to get all this generically working. So no matter how much I like labwc it will not replace my X wm choices. If my coding skills were anywhere near enough to rewrite vtwm I would, but you can't teach an old dog new code. Next life maybe.
Re: X11 vs Wayland: what does the timeline look like?
> At this time it looks like I cannot switch to Wayland -- several things > about Wayland just don't work (for me). > > I know of at least three things where Wayland simply fails: > > I use KiCAD and KiCAD just does not work with Wayland: > https://www.kicad.org/blog/2025/06/KiCad-and-Wayland-Support/ I haven't run anything that runs on X that will not run with Xwayland. > I use FVWM2 as my window manager. And I believe FVWM2 uses features of X11 > that are not available in Wayland. And no I don't like Gnome or KDE (or > really any so-called "modern" desktop environment. There is an α or β experimental still system called wayback, basically uses xwayland to provide an X environment to run an X wm or dt on top of wlroots. Yes, fvwm3 was one of the few I tried. There is also labwc which is an openbox port to wayland and it is a delight to play with. It is also about 1/10 the size of fvwm > Also I use SSH X11 tunneling very extensively. I have a whole LAN full of > little Linux machines (mostly assorted 'Pis). Some of these are lower end > machines that I don't want to run full-fledged desktop environments on and use > RDP or VNC with, expecially just to run a simple X11 program on. Cage is another low resource way to run an application alone is wayland, no wm needed. > So, is there a likely timeline where X11 (in whatever form that takes) > "vanishes" and Wayland totally replaces it as the graphical server > infratructure under Linux? Are we talking a few years, a decade, or several > decades? How would x11 vanish, if any of us have the code and a reliable place to store it there will always be fork around. Will C glibc/gcc change so much nobody will be able to update it so it will compile, even if, the kept binaries will last some centuries on the right antique machine. The reason I don't use it is 1 sudo -u or doas xxx-user X or Wayland app. refuses to start, no seat no xdg runtime and even when you cheat and get it able it would be isolated from the rest of the system as if it was a container. What I believe the problem is that it is an intentional decision not to allow user1 to run an app. as user2
Re: evince crashes the X server
On Thursday, July 10th, 2025 at 5:17 AM, Eyal Lebedinsky wrote: > On 10/7/25 14:57, Vladimir Dergachev wrote: > > > [163563.414] (EE) Backtrace: > > > [163563.416] (EE) unw_get_proc_name failed: no unwind info found [-10] > > > [163563.416] (EE) 0: /usr/lib64/xorg/modules/drivers/intel_drv.so (?+0x0) > > > [0x7f3b716deede] > > > [163563.416] (EE) unw_get_proc_name failed: no unwind info found [-10] > > > [163563.416] (EE) 1: /usr/lib64/xorg/modules/drivers/intel_drv.so (?+0x0) > > > [0x7f3b716d0a35] > > > [163563.416] (EE) unw_get_proc_name failed: no unwind info found [-10] either libunwind is not installed or it is a messed up pkg or non-compatible version installed? that's the first place I would check
Re: Question about the future of Xorg
> On Fri, 4 Jul 2025, Felipe Contreras wrote: > > On Tue, Jun 24, 2025 at 1:23 AM Michel Dänzer mic...@daenzer.net wrote: > > > On 2025-06-11 03:53, Felipe Contreras wrote: > > > > On Tue, Jun 10, 2025 at 1:11 AM Carsten Haitzler > > > > ras...@rasterman.com wrote: > > No, I'm not. Asking a question doesn't entail any expectation. When consolekit was abandoned, it was forked and continues today as consolekit2 When udev as independent project was abandoned it continued forked as eudev abandoned again forked and continued So now we have X server anticipated to be abandoned by some, not officially, but maybe a lesson was learned. If you were to present the fork before the fact, make adequate noise, create some momentum, you can control the fork as well, making a fork of a fork virtually obscure and invisible. This is what happens when you don't answer a question, conspiracy theories blow up to hollywood scenarios. Thanks to the BSD community though, we have little to fear ... corporate linux can only reflect pathogeny of behind the scenes trickery.
Re: [ANNOUNCE] igt-gpu-tools 2.1
Kamil Konieczny wrote: > A new igt-gpu-tools release is available with the following changes: > > Release 2.01 (2025-06-27) > - Should this read 2.1 instead v2.1 is in the source v2.01 is not
