x11vnc README file Date: Wed Dec 22 23:18:56 EST 2004 The following information is taken from these URLs: http://www.karlrunge.com/x11vnc/index.html http://www.karlrunge.com/x11vnc/x11vnc_opts.html they contain the most up to date info. ======================================================================= http://www.karlrunge.com/x11vnc/index.html: _________________________________________________________________ x11vnc: a VNC server for real X displays (to [1]FAQ) (to [2]downloads) (to [3]building) x11vnc allows one to remotely view and interact with real X displays (i.e. a display corresponding to a physical monitor, keyboard, and mouse) with any VNC viewer. In this way it plays the role for Unix/X11 that WinVNC plays for Windows. I wrote x11vnc because x0rfbserver was basically impossible to build on Solaris and had poor performance. The primary x0rfbserver build problems centered around esoteric C++ toolkits. x11vnc is written in plain C and uses only standard libraries. I also added a few enhancements to improve the interactive response, add esoteric features, etc. The [4]FAQ contains a lot of information and solutions to problems, but please feel free to [5]contact me if you have problems or questions. Background: VNC (Virtual Network Computing) is a very useful network graphics protocol in the spirit of X, however, unlike X, the viewing-end is very simple and maintains no state. It is a remote framebuffer (RFB) protocol Some VNC links: * [6]http://www.uk.research.att.com/vnc/ * [7]http://www.realvnc.com * [8]http://www.tightvnc.com For Unix, the VNC implementation includes a virtual X11 server Xvnc (usually launched via the vncserver command) that is not associated with a real display, but provides a "fake" one X11 clients (xterm, mozilla, etc.) can attach to. A remote user then connects to Xvnc via the VNC client vncviewer from anywhere on the network to view and interact with the whole virtual X11 desktop. The VNC protocol is in most cases better suited for remote connections with low bandwidth and high latency than is the X11 protocol. Also, with no state maintained the viewing-end can crash, be rebooted, or relocated and the applications and desktop continue running. Not so with X11. So the standard Xvnc program is very useful, I use it for things like: * desktop conferencing with other users (e.g. codereviews). * long running apps/tasks I want to be able to view from many places. * Motif, GNOME, and similar applications that would yield very poor performance over a high latency link. However, sometimes one wants to connect to a real X11 display (i.e. one attached to a physical monitor, keyboard, and mouse: a Workstation or a SunRay session) from far away. Maybe you want to close down an application cleanly rather than using kill, or want to work a bit in an already running application, or would like to help a distant colleague solve a problem with their desktop. This is where x11vnc is useful. _________________________________________________________________ How to use x11vnc: In this example let's assume the remote machine with the X display you wish to view is "far-away.east:0" and the workstation you are presently working at is "sitting-here.west". Step 0. Download x11vnc ([9]see below) and have it available to run (e.g. via $PATH) on far-away.east. Similarly, have a VNC viewer (e.g. vncviewer) ready to run on sitting-here.west. We recommend [10]TightVNC Viewers. Step 1. By some means log in to far-away.east and get a command shell running there. You can use ssh, rlogin, telnet, or any other method to do this. x11vnc needs to be run on the same machine the X server process is running on (because MIT-SHM shared memory is used to poll the X11 framebuffer). Step 2. In that far-away.east shell (with command prompt "far-away>" in this example) run x11vnc directed at the far-away.east X session display: far-away> x11vnc -display :0 You could have also set the environment variable DISPLAY=:0 instead of using -display. This step attaches x11vnc to the far-away.east:0 X display (no viewer clients yet). To get X11 permissions right, you may also need to set the XAUTHORITY environment variable (or use the -auth option) to point to the correct MIT-MAGIC-COOKIE file (e.g. /home/joe/.Xauthority). More on this [11]below. There will then be much chatter printed out from x11vnc, until it finally says something like: . . 13/05/2004 14:59:54 Autoprobing selected port 5900 13/05/2004 14:59:54 screen setup finished. 13/05/2004 14:59:54 The VNC desktop is far-away:0 PORT=5900 which means all is OK, and we are ready for the final step. Step 3. At the place where you are sitting (sitting-here.west in this example) you now want to run a VNC viewer program. There are VNC viewers for Unix, Windows, MacOS, Java-enabled web browsers, and even for PDA's like the Palm Pilot! You can use any of them to connect to x11vnc (see the above VNC links under "Background:" on how to obtain a viewer for your platform or [12]this FAQ. For Solaris, vncviewer is available in the [13]Companion CD package SFWvnc ). In this example we'll use the Unix vncviewer program on sitting-here by typing the following command in a second terminal window: sitting-here> vncviewer far-away.east:0 That should pop up a viewer window on sitting-here.west showing and allowing interaction with the far-away.east:0 X11 desktop. Pretty nifty! When finished, exit the viewer: the remote x11vnc process will shutdown automatically (or you can use the -forever [14]option to have it wait for additional viewer connections). Desktop Sharing: The above more or less assumed nobody was sitting at the workstation display "far-away.east:0". This is often the case: a user wants to access her workstation remotely. Another usage pattern has the user sitting at "far-away.east:0" and invites one or more other people to view and interact with his desktop. Perhaps the user gives a demo or presentation this way (using the telephone for vocal communication). A "Remote Help Desk" mode would be similar: a technician remotely connects to the user's desktop to interactively solve a problem the user is having. For these cases it should be obvious how it is done. The above steps will work, but more easily the user sitting at far-away.east:0 simply starts up x11vnc from a terminal window, after which the guests would start their VNC viewers. For this usage mode the -accept popup option discussed in the [15]FAQ below may be of use to allow the user at far-away.east:0 to accept or reject incoming connections. _________________________________________________________________ Tunnelling x11vnc via ssh: The above example had no security or privacy at all. When logging into remote machines (certainly when going over the internet) it is best to use ssh, or use a VPN. For x11vnc one can tunnel the VNC protocol through the encrypted ssh channel. It would look something like this: sitting-here> ssh -L 5900:localhost:5900 far-away.east 'x11vnc -display :0' (you will likely have to provide passwords/passphrases for the ssh login) and then in another terminal window on sitting-here run the command: sitting-here> vncviewer -encodings "copyrect tight zrle hextile" localhost:0 The -encodings option is very important: vncviewer will default to "raw" encoding if it thinks the connection is to the local machine, and so vncviewer gets tricked this way by the ssh redirection. "raw" encoding will be extremely slow over a networked link, so you need to force the issue with -encodings "copyrect tight ...". If the machine you SSH into is not the same machine with the X display you wish to view (e.g. your company provides incoming SSH access to a gateway machine), then you need to change the above to, e.g.: -L 5900:otherhost:5900. Once logged in, you'll need to do a second login (ssh, rsh, etc.) to the workstation machine 'otherhost' and then start up x11vnc on it. Scripts to automate tunneling: As discussed below, there may be some problems with port 5900 being available. If that happens, the above port and display numbers may change a bit (e.g. -> 5901 and :1). However, if you "know" port 5900 will be free on the local and remote machines, you can easily automate the above two steps by using the x11vnc option -bg (forks into background after connection to the display is set up) or using the -f option of ssh. A simple example script, assuming no problems with port 5900 being taken on the local or remote sides, looks like: #!/bin/sh # usage: x11vnc_ssh : # e.g.: x11vnc_ssh snoopy.peanuts.com:0 host=`echo $1 | awk -F: '{print $1}'` disp=`echo $1 | awk -F: '{print $2}'` if [ "x$disp" = "x" ]; then disp=0; fi cmd="x11vnc -display :$disp -localhost -rfbauth .vnc/passwd" enc="copyrect tight zrle hextile zlib corre rre raw" ssh -f -L 5900:localhost:5900 $host "$cmd" for i in 1 2 3 do sleep 2 if vncviewer -encodings "$enc" :0; then break; fi done See also rx11vnc.pl below. Another method is to start the VNC viewer in listen mode "vncviewer -listen" and have x11vnc initiate a reverse connection using the -connect option: #!/bin/sh # usage: x11vnc_ssh : # e.g.: x11vnc_ssh snoopy.peanuts.com:0 host=`echo $1 | awk -F: '{print $1}'` disp=`echo $1 | awk -F: '{print $2}'` if [ "x$disp" = "x" ]; then disp=0; fi cmd="x11vnc -display :$disp -localhost -connect localhost" # <-- note new opt ion enc="copyrect tight zrle hextile zlib corre rre raw" vncviewer -encodings "$enc" -listen & pid=$! ssh -R 5500:localhost:5500 $host "$cmd" kill $pid A third way is specific to the TightVNC vncviewer special option -via for gateways. The only tricky part is we need to start up x11vnc and give it some time to start listening for connections (so we cannot use the TightVNC default setting for VNC_VIA_CMD): #!/bin/sh # usage: x11vnc_ssh : # e.g.: x11vnc_ssh snoopy.peanuts.com:0 host=`echo $1 | awk -F: '{print $1}'` disp=`echo $1 | awk -F: '{print $2}'` if [ "x$disp" = "x" ]; then disp=0; fi VNC_VIA_CMD="ssh -f -L %L:%H:%R %G x11vnc -localhost -rfbport 5900 -display :$d isp; sleep 5" export VNC_VIA_CMD vncviewer -via $host localhost:0 # must be TightVNC vncviewer. Of course if you already have the x11vnc running waiting for connections (or have it started out of inetd(1)), you can simply use the TightVNC vncviewer -via gateway host:port in its default mode to provide secure ssh tunnelling. VNC password file: Also note in the first example script that the option "-rfbauth .vnc/passwd" provides additional protection by requiring a VNC password for every VNC viewer that connects. The vncpasswd or storepasswd programs, or the x11vnc [16]-storepasswd option can be used to create the password file. x11vnc also has the slightly less secure [17]-passwdfile and "-passwd XXXXX" options. Important: It is up to you to tell x11vnc to use password protection, it will not do it for you automatically. The same goes for encrypting the channel between the viewer and x11vnc: it is up to you to use ssh, stunnel, VPN, etc. Also look into the -allow and -localhost [18]options and building x11vnc with [19]tcp_wrappers support to limit host access. _________________________________________________________________ Downloading x11vnc: x11vnc is a contributed program to the [20]libvncserver project at SourceForge.net. I use libvncserver for all of the VNC aspects; I couldn't have done without it. The full source code may be found and downloaded (either file-release tarball or CVS tree) from the above link. As of Aug 2004, the [21]x11vnc-0.6.2.tar.gz source package is released (recommended download) . The x11vnc package is the subset of the libvncserver package needed to build the x11vnc program. Please do not use the LibVNCServer-0.6 tarball: it contains an older, more buggy version of x11vnc (Oct 2003) that you likely want to avoid. Also, you can get a copy of my latest, bleeding edge [22]x11vnc.c file to replace the one in the above packages or the one in the CVS tree and then rebuild. If you have an older libvncserver source tree, you may need to switch on the OLD_TREE near the top of the x11vnc.c file. See the [23]FAQ below for information about where you might obtain a precompiled x11vnc binary from 3rd parties. To obtain VNC viewers for the viewing side (Windows, Mac OS, or Unix) try here: * [24]http://www.tightvnc.com/download.html * [25]http://www.realvnc.com/download-free.html * [26]http://sourceforge.net/projects/cotvnc/ More tools: Here is a rsh/ssh wrapper script rx11vnc that attempts to automatically do the above Steps 1-3 for you (provided you have rsh/ssh login permission on the machine x11vnc is to be run on). The above example would be: rx11vnc far-away.east:0 typed into a shell on sitting-here.west. Also included is an experimental script rx11vnc.pl that attempts to tunnel the vnc traffic through an ssh port redirection (and does not assume port 5900 is free). Have a look at them to see what they do and customize as needed: * [27]rx11vnc wrapper script * [28]rx11vnc.pl wrapper script to tunnel traffic thru ssh _________________________________________________________________ Building x11vnc: If your OS has libjpeg.so and libz.so in standard locations you can build as follows (example given for the 0.6.2 release of x11vnc: replace with the version you downloaded): (un-tar the x11vnc+libvncserver tarball) # gzip -dc x11vnc-0.6.2.tar.gz | tar -xvf - (cd to the source directory) # cd x11vnc-0.6.2 (run configure and then run make) # ./configure # make (if all went OK, copy x11vnc to the desired destination, e.g. $HOME/bin) # cp ./x11vnc/x11vnc $HOME/bin Or do make install, it will probably install to /usr/local/bin (run ./configure --help for information on customizing your configuration). You can now run it via typing "x11vnc", "x11vnc -help", "x11vnc -nap -display :0", etc. Note: Currently gcc is required to build libvncserver. In some cases it will build with non-gcc compilers, but the resulting binary often fails to run properly. Building on Solaris: Depending on your version of Solaris (or other Unix OS), the jpeg and/or zlib libraries may be in non-standard places (e.g. /usr/local, /usr/sfw, /opt/sfw, etc). Note: If configure cannot find these two libraries then TightVNC and ZRLE encoding support will be disabled, and you don't want that! (the TightVNC encoding gives very good compression and performance, it even makes a noticable difference over a fast LAN) libjpeg is included in Solaris 9 and later (/usr/sfw/include and /usr/sfw/lib), and zlib in Solaris 8 and later (/usr/include and /usr/lib). To get the source for these libraries: libjpeg is available at [29]ftp://ftp.uu.net/graphics/jpeg/ and zlib at [30]http://www.gzip.org/zlib/. See also [31]http://www.sunfreeware.com/ for Solaris binary packages of these libraries. Here is a build script that indicates one way to pass the library locations information to the libvncserver configuration: #!/bin/sh # Build script for Solaris, etc, with gcc, libjpeg and libz in # non-standard locations. PATH=/path/to/gcc/bin:/usr/ccs/bin:$PATH # set to get gcc JPEG=/path/to/jpeg # maybe "/usr/local", "/usr/sfw", or "/opt/sfw" ZLIB=/path/to/zlib # maybe "/usr/local", "/usr/sfw", or "/opt/sfw" # Below we assume headers in $JPEG/include and $ZLIB/include and the # shared libraries are in $JPEG/lib and $ZLIB/lib. If your situation # is different change the locations in the two lines below. CPPFLAGS="-I $JPEG/include -I $ZLIB/include" LDFLAGS="-L $JPEG/lib -R $JPEG/lib -L $ZLIB/lib -R $ZLIB/lib" # These two lines may not be needed on more recent Solaris releases: CPPFLAGS="$CPPFLAGS -I /usr/openwin/include" LDFLAGS="$LDFLAGS -L /usr/openwin/lib -R /usr/openwin/lib" # This only applies to Solaris 10 or later (XFIXES and DAMAGE) LDFLAGS="$LDFLAGS -L /usr/openwin/sfw/lib -R /usr/openwin/sfw/lib" # Everything needs to built with _REENTRANT for thread safe errno: CPPFLAGS="$CPPFLAGS -D_REENTRANT" export PATH CPPFLAGS LDFLAGS ./configure make ls -l ./x11vnc/x11vnc Then do make install or copy the binary to your desired destination. Note that on Solaris make is /usr/ccs/bin/make, so that is why the above puts /usr/ccs/bin in PATH. Other important build utililties are there too: ld, ar, etc. Also, it is probably a bad idea to have /usr/ucb in your PATH while building. One user had a problem where the above build script was failing because his work environment had the ENV variable set to a script that was resetting his PATH so that gcc could no longer be found. Make sure you do not have any ENV or BASH_ENV in your environment doing things like that. There is a build problem on Solaris 7 11/99 (update 4) where the header file X11/extensions/XKBstr.h that X11/XKBlib.h uses was not shipped. The x11vnc configure succeeds and sets LIBVNCSERVER_HAVE_XKEYBOARD in rfb/rfbconfig.h but then the build of x11vnc fails in the make. A workaround is to remove all lines referring to LIBVNCSERVER_HAVE_XKEYBOARD in rfb/rfbconfig.h after configure has been run. Alternatively, one could put #undef LIBVNCSERVER_HAVE_XKEYBOARD after the rfb/rfb.h include in the x11vnc/x11vnc.c file. (This problem has been fixed as of x11vnc 0.6.2 (Aug/2004)) If you need to build on Solaris 2.5.1 or earlier, see [32]this workaround FAQ. Building on HP-UX: For jpeg and zlib you will need to do the same sort of thing as described above for Solaris. You set CPPFLAGS and LDFLAGS to find them. You do not need to do any of the above /usr/openwin stuff. Also, HP-UX does not seem to support -R, so get rid of the -R items in LDFLAGS. Because of this, at runtime you may need to set LD_LIBRARY_PATH to indicate the directory paths so the libraries can be found. Finally, there seems to be a bug with gcc on HP-UX 11.xx: something fails (in the gcc private header files?) and it thinks it cannot find gettimeofday(). As a workaround add this to CPPFLAGS: -DLIBVNCSERVER_HAVE_GETTIMEOFDAY. You may get some warnings but we have verified that this generates working x11vnc binaries on HP-UX hppa and ia64. _________________________________________________________________ Some Notes: Network performance: Whether you are using Xvnc or x11vnc it is always a good idea to have a solid background color instead of a pretty background image. Each and every re-exposure of the background must be resent over the network: better to have that background be a solid color that compresses very well compared to a photo image. (This is one place where the X protocol has an advantage over the VNC protocol.) I suggest using xsetroot, dtstyle or similar utility to set a solid background while using x11vnc. You can turn the pretty background image back on when you are using the display directly. I also find the [33]tightvnc encoding gives the best response for my usage (Unix <-> Unix over cable modem). One needs a tightvnc-aware vncviewer to take advantage of this encoding. TCP port issues: Notice the lines 18/07/2003 14:36:31 Autoprobing selected port 5900 PORT=5900 in the output. 5900 is the default VNC listening port (just like 6000 is X11's default listening port). Had port 5900 been taken by some other application, x11vnc would have next tried 5901. That would mean the viewer command above should be changed to vncviewer far-away.east:1. You can force the issue with the "-rfbport NNNN" option where NNNN is the desired port number. Options: x11vnc has (far too) many features that may be activated via its [34]command line options. Useful options are -nap to use fewer resources (it sleeps more between polls when activity is low) and -rfbauth passwd-file to use VNC password protection (the vncpasswd or storepasswd programs, or the x11vnc [35]-storepasswd option can be used to create the password file). Algorithm: How does x11vnc do it? Rather brute-forcedly: it continuously polls the X11 framebuffer for changes using XShmGetImage(). When changes are discovered, it instructs libvncserver which rectangular regions of the framebuffer have changed, and libvncserver compresses the changes and sends them off to any connected VNC viewers. A number of applications do similar things, such as x0rfbserver, krfb, x0vncserver. x11vnc uses a 32 x 32 pixel tile model (the desktop is decomposed into ~1000 such tiles), where changed tiles are found by pseudo-randomly polling 1 pixel tall horizontal scanlines. This is a surprisingly effective algorithm for finding changed regions. For keyboard and mouse user input the XTEST extension is used to pass the input events to the X server. To detect XBell "beeps" the XKEYBOARD extension is used. If available, the XFIXES extension is used to retrieve the current mouse cursor shape. Barbershop mirrors effect: What if x11vnc is started up, and vncviewer is then started up on the same machine and displayed on the same display x11vnc is polling? One might "accidentally" do this when first testing out the programs. You get an interesting "feedback" effect where vncviewer images keep popping up each one contained in the previous one and slightly shifted a bit by the window manager decorations. There will be an [36]even more interesting effect if -scale is used. Also, if the XKEYBOARD is supported and the XBell "beeps" once, you get an infinite loop of beeps going off. Although all of this is mildly exciting it is not much use: you will normally run and display the viewer on a different machine! SunRay notes: You can run x11vnc on your (connected or disconnected) [37]SunRay session (Please remember to use -nap and maybe -wait 200 to avoid being a resource hog! It also helps a bit to have a solid background color). You have to know the name of the machine your SunRay session X server is running on. You also need to know the X11 DISPLAY number for the session: on a SunRay it could be a large number, e.g. :137, since there are many people with X sessions (Xsun processes) on the same machine. If you don't know it, you can get it by running who(1) in a shell on the SunRay server and looking for the dtlocal entry with your username (and if you don't even know which server machine has your session, you could login to all possible ones looking at the who output for your username...). SunRay Gotcha #1: Note that even though your SunRay X11 DISPLAY is something like :137, x11vnc still tries for port 5900 as its listening if it can get it, in which case the VNC display (i.e. the information you supply to the VNC viewer) is something like sunray-server:0 (note the :0 corresponding to port 5900, it is not :137). If it cannot get 5900, it tries for 5901, and so on. You can also try to force the port (and thereby the VNC display) using the -rfbport NNNN option. Limitations: * Due to the polling nature, some activities (opaque window moves, scrolling), can be pretty choppy/ragged and others (exposures of large areas) slow. Experiment with interacting a bit differently than you normally do to minimize the effects (e.g. do fullpage paging rather than line-by-line scrolling, and move windows in a single, quick motion). Work is in progress to attempt to speed things up using the copyrect encoding and other things, but they will likely only speed up certain activities, not all. * A rate limiting factor for x11vnc performance is that video hardware is optimized for writing, not reading (x11vnc reads the video framebuffer for the screen image data). The difference can be a factor of 10-50, and it usually takes about 0.5-1 sec to read in the whole video hardware framebuffer (5MB for 1280x1024 at depth 24). So whenever activity changes most of the screen there is a delay of this length while x11vnc reads the changed regions in. To get a sense of the read and write speeds of your video card, you can run the benchmarks like: x11perf -getimage500, x11perf -putimage500, x11perf -shmput500 and for XFree86 displays with direct graphics access the dga command (press "b" to run the benchmark and then after a few seconds press "q" to quit). On XFree86 it is actually possible to increase the framebuffer read speed considerably (5-50 times) by using the Shadow Framebuffer (a copy of the framebuffer is kept in main memory and this can be read much more quickly). To do this one puts the line Option "ShadowFB" "true" (and depending on video card driver, Option "NoAccel" "true" may be needed too) in the Device section of the /etc/X11/XF86Config file. Note that this disables 2D acceleration at the physical display and so likely defeats the purpose. Nevertheless this could be handy in some circumstances, e.g. if the slower speed at the physical display was acceptable (this seems to be true for most video cards these days). Unfortunately it does not seem shadowfb can be turned on and off dynamically... * Somewhat surprisingly, the X11 mouse (cursor) shape is write-only and cannot be queried from the X server. So in x11vnc the cursor shape stays fixed at an arrow. (see the "-cursor X" and "-cursor some", however, for a partial hack for the root window, etc.). Also, on Solaris using the SUN_OVL overlay extension, x11vnc can show the correct mouse cursor when the -overlay option is also supplied. A similar thing is done on IRIX as well when -overlay is supplied. As of Dec/2004 x11vnc supports the XFIXES extension (in Xorg and Solaris 10) to query the X server for the exact cursor shape, this works pretty well. * Audio from applications is of course not redirected (separate redirectors do exist, e.g. esd). The XBell() "beeps" will work if the X server supports the XKEYBOARD extension. (Note that on Solaris XKEYBOARD is disabled by default. Passing +kb to Xsun enables it). * Occasionally a patch of tiles will not get updated correctly. Evidently a timing related bug and difficult to reproduce... * Using -threads can expose some bugs in libvncserver. Please feel free to [38]contact me if you have any questions, problems, or comments about x11vnc, etc. _________________________________________________________________ x11vnc FAQ: [39]Q-1: I can't get x11vnc to start up. It says "XOpenDisplay failed ((null))" or "Xlib: connection to ":0.0" refused by server". What do I need to do? [40]Q-2: I can't get x11vnc and/or libvncserver to compile. [41]Q-3: Help, I need to run x11vnc on Solaris 2.5.1 and it doesn't compile! If I try to run a binary built on Solaris 2.6 I get: relocation error: file x11vnc: symbol XConvertCase: referenced symbol not found [42]Q-4: Where can I get a precompiled x11vnc binary for my Operating System? [43]Q-5: Where can I get a VNC Viewer binary (or source code) for the Operating System I will be viewing from? [44]Q-6: How can I see all of x11vnc's command line options and documentation on how to use them? [45]Q-7: I don't like typing arcane command line options every time I start x11vnc. What can I do? Is there a config file? Is there a GUI? [46]Q-8: Why does x11vnc exit as soon as the VNC viewer disconnects? And why doesn't it allow more than one VNC viewer to connect at the same time? [47]Q-9: I have two separate machine displays in front of me, one Windows the other X11: can I use x11vnc in combination with Win2VNC in dual-screen mode to pass the keystrokes and mouse motions to the X11 display? [48]Q-10: I am running Win2VNC on my windows machine and trying to create a dual-screen mode with my second display by running "x11vnc -nofb". Whenever I initiate the connection Win2VNC quickly disconnects and x11vnc says something like: rfbProcessClientNormalMessage: read: Connection reset by peer [49]Q-11: The X display I run x11vnc on is only 8 bits per pixel (bpp) PseudoColor (i.e. only 256 distinct colors). The x11vnc colors may start out OK, but after a while the colors are incorrect in certain windows. [50]Q-12: Color problems: Why are the colors for some windows messed up in x11vnc? BTW, I have an X display that has nice overlay/multi-depth visuals of different color depths: e.g. there are both depth 8 and 24 visuals available at the same time. [51]Q-13: How do I figure out the window id to supply to the -id windowid option? [52]Q-14: Why don't menus or other transient windows come up when I am using the -id windowid option to view a single application window? [53]Q-15: Can I use x11vnc to view and interact with an Xterminal (e.g. NCD) that is not running UNIX and so x11vnc cannot be run on it directly? [54]Q-16: Can I make x11vnc more quiet and also go into the background after starting up? [55]Q-17: How do I stop x11vnc once it is running in the background? [56]Q-18: Can I change settings in x11vnc without having to restart it? Is there a way to remote control it? [57]Q-19: Can I limit which machines incoming VNC clients can connect from? [58]Q-20: How do I build x11vnc/libvncserver with libwrap (tcp_wrappers) support? [59]Q-21: Can I prompt the user at the local X display whether the incoming VNC client should be accepted or not? Can I decide to make some clients view-only? How about running an arbitrary program to make the decisions? [60]Q-22: How do I create a VNC password for use with x11vnc? [61]Q-23: How can I tunnel my connection to x11vnc via an encrypted SSH channel between two Unix machines? [62]Q-24: How can I tunnel my connection to x11vnc via an encrypted SSH channel from Windows using an SSH client like Putty? [63]Q-25: Does x11vnc support Unix usernames and passwords? Can I further limit the set of Unix usernames who can connect to the VNC desktop? [64]Q-26: Can I have two passwords for VNC viewers, one for full access and the other for view-only access to the display? [65]Q-27: I use a screen-lock when I leave my workstation (e.g. xscreensaver or xlock). When I remotely access my workstation desktop via x11vnc I can unlock the desktop fine, but I am worried people will see my activities on the physical monitor. What can I do to prevent this, or at least make it more difficult? [66]Q-28: Can I have x11vnc automatically lock the screen when I disconnect the VNC viewer? [67]Q-29: Are reverse connections (i.e. the VNC server connecting to the VNC viewer) using "vncviewer -listen" and vncconnect(1) supported? [68]Q-30: Sometimes when a VNC viewer dies abruptly, x11vnc also dies with the error message like: "Broken pipe". I'm using the -forever mode and I want x11vnc to keep running. [69]Q-31: How can I use x11vnc to connect to an X login screen like xdm, GNOME gdm, KDE kdm, or CDE dtlogin? (i.e. nobody is logged into an X session yet). [70]Q-32: Can I run x11vnc out of inetd(1)? [71]Q-33: How do I make x11vnc work with the Java VNC viewer applet in a web browser? [72]Q-34: Why isn't the mouse cursor shape (the little icon shape where the mouse pointer is) correct as I move from window to window? [73]Q-35: Why does the mouse arrow just stay in one corner in my vncviewer, whereas my cursor (that does move) is just a dot? [74]Q-36: Can I take advantage of the TightVNC extension to the VNC protocol where Cursor Positions Updates are sent back to all connected clients (i.e. passive viewers can see the mouse cursor being moved around by another viewer)? [75]Q-37: Is it possible to swap the mouse buttons (e.g. left-handed operation), or arbitrarily remap them? How about mapping button clicks to keystrokes, e.g. to partially emulate Mouse wheel scrolling? [76]Q-38: When I drag windows around with the mouse or scroll up and down things really bog down (unless I do the drag in a single, quick motion). Is there anything to do to improve things? [77]Q-39: I have lots of memory, but why does x11vnc fail with shmget: No space left on device or Minor opcode of failed request: 1 (X_ShmAttach)? [78]Q-40: How can I make x11vnc use less system resources? [79]Q-41: How can I make x11vnc use MORE system resources? [80]Q-42: I use x11vnc over a slow link with high latency (e.g. dialup modem), is there anything I can do to speed things up? [81]Q-43: How can I get my AltGr and Shift modifiers to work between keyboards for different languages? [82]Q-44: When I try to type a "<" (i.e. less than) instead I get ">" (i.e. greater than)! Strangely, typing ">" works OK!! [83]Q-45: I'm using an "international" keyboard (e.g. German "de", or Danish "dk") and the -modtweak mode works well if the VNC viewer is run on a Unix/Linux machine with a similar keyboard. But if I run the VNC viewer on Unix/Linux with a different keyboard (e.g. "us") or Windows with any keyboard, I can't type some keys like: "@", "$", "<", ">", etc. How can I fix this? [84]Q-46: When typing I sometimes get double, triple, or more of my keystrokes repeated. I'm sure I only typed them once, what can I do? [85]Q-47: The machine where I run x11vnc has an AltGr key, but the local machine where I run the VNC viewer does not. Is there a way I can map a local unused key to send an AltGr? How about a Compose key as well? [86]Q-48: I have a Sun machine I run x11vnc on. Its Sun keyboard has just one Alt key labelled "Alt" and two Meta keys labelled with little diamonds. The machine where I run the VNC viewer only has Alt keys. How can I send a Meta keypress? (e.g. emacs needs this) [87]Q-49: Can I map a keystroke to a mouse button click on the remote machine? [88]Q-50: The remote display is larger (in number of pixels) than the local display I am running the vncviewer on. I don't like the vncviewer scrollbars, what I can do? [89]Q-51: Does x11vnc support server-side framebuffer scaling? (E.g. to make the desktop smaller). [90]Q-52: Does x11vnc work with Xinerama? (i.e. multiple monitors joined together to form one big, single screen). [91]Q-53: Can I use x11vnc on a multi-headed display that is not Xinerama (i.e. separate screens :0.0, :0.1, ... for each monitor)? [92]Q-54: Does x11vnc support the XRANDR (X Resize, Rotate and Reflection) extension? Whenever I rotate or resize the screen x11vnc just seems to crash. [93]Q-55: Why is the view in my VNC viewer completely black? Or why is everything flashing around randomly? [94]Q-56: I use Linux Virtual Consoles (VC's) to implement 'Fast User Switching' between users' sessions (e.g. Betty is on Ctrl-Alt-F7, Bobby is on Ctrl-Alt-F8, and Sid is on Ctrl-Alt-F1: they use those keystrokes to switch between their sessions). How come the view in a VNC viewer connecting to x11vnc is either completely black or otherwise all messed up unless the X session x11vnc is attached to is in the active VC? [95]Q-57: Does the Clipboard/Selection get transferred between the vncviewer and the X display? [96]Q-58: Why don't I hear the "Beeps" in my X session (e.g. when typing tput bel in an xterm)? [97]Q-59: I am using x11vnc where my local machine has "popup/hidden taskbars" (e.g. GNOME or MacOS X) and the remote display where x11vnc runs also has "popup/hidden taskbars" (e.g. GNOME). When I move the mouse to the edge of the screen where the popups happen, the taskbars interfere and fight with each other in strange ways. What can I do? _________________________________________________________________ Q-1: I can't get x11vnc to start up. It says "XOpenDisplay failed ((null))" or "Xlib: connection to ":0.0" refused by server". What do I need to do? For the former error, you need to specify the X display to connect to (it also needs to be on the same machine x11vnc is to run on). Set your DISPLAY environment variable or use the -display option to specify it. Nearly always the correct value will be ":0" For the latter error, you need to set up the X11 permissions correctly. See the xauth(1), Xsecurity(7), and xhost(1) man pages for much info. For example, you may need to set your XAUTHORITY environment variable (or use the -auth option) to point to the correct cookie file (e.g. /home/joe/.Xauthority or /var/lib/xdm/authdir/authfiles/A:0-nRySEi), or simply be sure you run x11vnc as the correct user (i.e. the user who owns the X session you wish to view). Less safe, but if the owner of the X session runs xhost +localhost one should be able to attach x11vnc to the session (from the same machine). We have been informed that some recent Linux distribution display managers set XAUTHORITY to random local filenames, e.g.: /tmp/.gdmHaVtYy. So logging in as the correct user may not be enough... you need to dig out where they have hidden the MIT-MAGIC-COOKIE. The -e option to ps(1) may help, e.g.: ps wwwweaux | tr ' ' '\n' | grep XAUTHORITY | sort -u Note: use of Display Manager (gdm, kdm, ...) auth cookie files (i.e. from /var/..., /tmp/..., or elsewhere) may require modification via xauth(1) to correctly include the display x11vnc refers to (e.g. xauth -f cookie.file add :0 . 45be51ae2ce9dfbacd882ab3ef8e96b1, where 45be51... was found from an xauth -f /var/... list) or other reasons. See xauth(1) manpage for full details on how to transfer an MIT-MAGIC-COOKIE between machines and displays. To test out your X11 permissions, set DISPLAY (and possibly XAUTHORITY) and type xclock in the same place you will be typing (or otherwise running) x11vnc. If a little clock comes up on the display, that means the X11 permissions are OK. To test your X11 permissions when logged into the display machine remotely, use xdpyinfo (if you see the informational output about the display you know it connected to the X server successfully). Important: if you cannot get your X11 permissions so that the xclock or xdpyinfo tests work, x11vnc also will not work (all of these X clients must be able to connect to the X server to function). Q-2: I can't get x11vnc and/or libvncserver to compile. Make sure you have all of the required -devel packages installed. These include X11/XFree86, libjpeg, libz, ... After running the libvncserver configure, carefully examine the output and the messages in the config.log file looking for missing components. If the configure output looks like: checking how to run the C preprocessor... gcc -E checking for X... no checking for XkbSelectEvents in -lX11... no checking for XineramaQueryScreens in -lXinerama... no checking for XTestFakeKeyEvent in -lXtst... no there is quite a bit wrong with the build environment. Hopefully simply adding -devel packages will fix it. For Debian the list seems to be: gcc make libc6-dev libjpeg62-dev libx11-dev libxext-dev libxtst-dev x-dev xlibs-static-dev zlib1g-dev For Redhat the list seems to be: gcc make glibc-devel libjpeg-devel XFree86-devel zlib-devel Q-3: Help, I need to run x11vnc on Solaris 2.5.1 and it doesn't compile! If I try to run a binary built on Solaris 2.6 I get: relocation error: file x11vnc: symbol XConvertCase: referenced symbol not found We apologize that x11vnc does not build cleanly on older versions of Solaris, Linux, etc.: very few users are on these old releases. Here is a workaround for Solaris 2.5.1 (and perhaps earlier): First use the environment settings (CPPFLAGS, LDFLAGS, etc.) in the above [98]Solaris build script to run the configure command. That should succeed without failure. Then, you have to hand edit the autogenerated rfb/rfbconfig.h file in the source tree, and just before the last #endif at the bottom of that file insert these workaround lines: #ifndef usleep struct timeval _tmp_usleep_tv; #define usleep(x) \ _tmp_usleep_tv.tv_sec = (x) / 1000000; \ _tmp_usleep_tv.tv_usec = (x) % 1000000; \ select(0, NULL, NULL, NULL, &_tmp_usleep_tv); #endif int gethostname(char *name, int namelen); long random(); int srandom(unsigned int seed); #undef LIBVNCSERVER_HAVE_LIBPTHREAD #define SHUT_RDWR 2 typedef unsigned int in_addr_t; #ifndef snprintf #define snprintf(a, n, args...) sprintf((a), ## args) #endif Then run make with the Solaris build script environment, everything should compile without problems, and the resulting x11vnc binary should work OK (but note the workaround for XConvertCase in x11vnc.c only covers Latin 1). If some non-x11vnc related programs fail (e.g. test programs) and the x11vnc binary is not created try "make -k". Similar sorts of kludges can be done on other older OS (Solaris, Linux, ...) releases. Please let us know if you had to use the above workaround (and whether it worked or not). If there is enough demand we will try to push clean compilations back to earlier Solaris, Linux, etc, releases. Q-4: Where can I get a precompiled x11vnc binary for my Operating System? Hopefully the [99]build steps above and [100]FAQ provide enough info for a painless compile for most environments. Please report problems with the x11vnc configure, make, etc. on your system (if your system is known to compile other GNU packages successfully). There are precompiled x11vnc binaries made by other groups available at the following locations: Debian: (.deb) [101]http://packages.debian.org/x11vnc Slackware: (.tgz) [102]http://www.linuxpackages.net/ Redhat/Fedora: (.rpm) [103]http://dag.wieers.com/packages/x11vnc/ wwexptools: (.tgz) [104]http://www.bell-labs.com/project/wwexptools/packages.html The last one, wwexptools, provides a variety of Unix binaries (Linux, Solaris, HP-UX, Irix, ...) with the intent of being compatible on a wide range of OS releases. Find x11vnc near the bottom of that page and select 'download info'. Note: it appears some of the wwexptools x11vnc binaries are not compiled with libz and libjpeg and so the fast compression encodings: "Tight", "ZRLE", etc are not supported by those binaries. You can see this by looking at the x11vnc output and if it says the encoding for a client is "hextile" then likely the fast compression encodings are missing. If you want optimal performance on your OS, you should see the [105]build notes above for where to download libz and libjpeg, and then build everything with gcc. If any of the above binaries don't work and building x11vnc on your OS fails (and all else fails!) you can contact me as I often have a test binary I could give you. As a general note, the x11vnc program is simple enough you don't really need to install a package: the binary will in most cases work as is and from any location (as long as your system libraries are not too old, etc). So, for Linux distributions that are not one of the above, the x11vnc binary from the above packages has a good chance of working. You can "install" it by just copying the x11vnc binary to the desired directory in your PATH. Tip on extracting files from a Debian package: extract the archive via a command like: "ar x x11vnc_0.6-2_i386.deb" and then you can find the binary in the resulting data.tar.gz tar file. Also, rpm2cpio(1) is useful in extracting files from rpm packages. Q-5: Where can I get a VNC Viewer binary (or source code) for the Operating System I will be viewing from? To obtain VNC viewers for the viewing side (Windows, Mac OS, or Unix) try here: * [106]http://www.tightvnc.com/download.html * [107]http://www.realvnc.com/download-free.html * [108]http://sourceforge.net/projects/cotvnc/ Q-6: How can I see all of x11vnc's command line options and documentation on how to use them? Run: x11vnc -help The output is listed [109]here as well. Q-7: I don't like typing arcane command line options every time I start x11vnc. What can I do? Is there a config file? Is there a GUI? You could create a shell script that calls x11vnc with your options: #!/bin/sh # # filename: X11vnc (i.e. not "x11vnc") # It resides in a directory in $PATH. "chmod 755 X11vnc" has been run on it. # x11vnc -nap -wait 50 -localhost -rfbauth $HOME/.vnc/passwd -display :0 $* a similar thing can be via aliases in your shell (bash, tcsh, csh, etc..). Or as of Jun/2004 in the libvncserver CVS you can use the simple $HOME/.x11vncrc config file support. If that file exists, each line is taken as a command line option. E.g. the above would be: # this is a comment in my ~/.x11vncrc file nap wait 50 # comment to the end of the line -localhost # note: the leading "-" is optional. rfbauth /home/fred/.vnc/passwd display :0 At some point this config file support will be expanded. As of Dec/2004 in the libvncserver CVS there is now a simple tcl/tk GUI based on the remote-control functionality that was added. It's not particulary user-friendly, it just provides a point and click mode to set all the many x11vnc parameters and obtain help on them. See the -gui for more info. Examples: "x11vnc ... -gui" and "x11vnc ... -gui other:0" in the latter case the gui is displayed on other:0 not the X display x11vnc is polling. Q-8: Why does x11vnc exit as soon as the VNC viewer disconnects? And why doesn't it allow more than one VNC viewer to connect at the same time? These defaults are simple safety measures to avoid someone unknowingly leaving his X11 desktop exposed (to the internet, say) for long periods of time. Use the -forever option (aka -many) to have x11vnc wait for more connections after the first client disconnects. Use the -shared option to have x11vnc allow multiple clients to connect simultaneously. Recommended additional safety measures include using ssh ([110]see above), stunnel, or a VPN to authenticate and encrypt the viewer connections or to at least use the -rfbauth passwd-file [111]option to use VNC password protection (or [112]-passwdfile) It is up to you to apply these security measures, they will not be done for you automatically. Q-9: I have two separate machine displays in front of me, one Windows the other X11: can I use x11vnc in combination with Win2VNC in dual-screen mode to pass the keystrokes and mouse motions to the X11 display? Yes, for best response start up x11vnc with the "-nofb" option (disables framebuffer polling, and does other optimizations) on the secondary display (X11). Then start up Win2VNC on the primary display (Windows) referring it to the secondary display. This will also work X11 to X11 using [113]x2vnc, however you would probably just want to avoid VNC and use x2x for that. For reference, here are some links to Win2VNC-like programs for multiple monitor setups: * [114]Original Win2VNC * [115]Enhanced Win2VNC and [116]sourceforge link * [117]x2vnc * [118]x2x also [119]here * [120]zvnc (MorphOS) All of them (except x2x) will work with x11vnc. Q-10: I am running Win2VNC on my windows machine and trying to create a dual-screen mode with my second display by running "x11vnc -nofb". Whenever I initiate the connection Win2VNC quickly disconnects and x11vnc says something like: rfbProcessClientNormalMessage: read: Connection reset by peer Is the default visual of the X display you run x11vnc on low color (e.g. 8 bit per pixel PseudoColor)? (you can run xdpyinfo to check). There seems to be a bug in Win2VNC in that it cannot deal correctly with colormaps (PseudoColor is the most common example of a visual with a colormap). If so, there are a couple options. 1) Can you set the default visual on your display to be depth 24 TrueColor? Sun machines often have 8+24 overlay/multi-depth visuals, and you can make the default visual depth 24 TrueColor (see fbconfig(1) and Xsun(1)). 2) As of Feb/2004, in the libvncserver CVS, x11vnc has the -visual option to allow you to force the framebuffer visual to whatever you want (this usually messes up the colors unless you are very careful). In this case, the option provides a convenient workaround for the Win2VNC bug: x11vnc -nofb -visual TrueColor -display :0 ... So the visual will be set to 8bpp TrueColor and Win2VNC can handle this. Since Win2VNC does not use the framebuffer data there should be no problems in doing this. Q-11: The X display I run x11vnc on is only 8 bits per pixel (bpp) PseudoColor (i.e. only 256 distinct colors). The x11vnc colors may start out OK, but after a while the colors are incorrect in certain windows. Use the -flashcmap option to have x11vnc watch for changes in the colormap, and propagate those changes back to connected clients. This can be slow (since the whole screen must be updated over the network whenever the colormap changes). This flashing colormap behavior often happens if an application installs its own private colormap when the mouse is in its window. "netscape -install" is a well-known historical example of this. Consider reconfiguring the system to 16 bpp or depth 24 TrueColor if at all possible. Also note that in some rare cases the -notruecolor option has corrected colors on 8bpp displays. The red, green, and blue masks were non-zero in 8bpp PseudoColor on an obscure setup, and this option corrected the problems. Q-12: Color problems: Why are the colors for some windows messed up in x11vnc? BTW, I have an X display that has nice overlay/multi-depth visuals of different color depths: e.g. there are both depth 8 and 24 visuals available at the same time. You may want to review the [121]previous question regarding 8 bpp PseudoColor. If that isn't the problem, run xdpyinfo(1) to see what the default visual is and what the depths of the other visuals are. Does the default visual have a depth of 8? If it does, can you possibly re-configure your X server to make the depth 24 visual the default? If you can do it, this will save you a lot of grief WRT colors and x11vnc (and for general usage too!). Here is how I do this on an old Sparcstation 20 running Solaris 9 with SX graphics xinit -- -dev /dev/fb defclass TrueColor defdepth 24 and it works nicely (note: to log into console from the dtlogin window, select "Options -> Command Line Login", then login and enter the above command). See the -dev section of the Xsun(1) manpage for a description of the above arguments. If you have root permission, a more permanent and convenient thing to do is to record the arguments in a line like: :0 Local local_uid@console root /usr/openwin/bin/Xsun -dev /dev/fb defclass TrueColor defdepth 24 in /etc/dt/config/Xservers (see /usr/dt/config/Xservers). Also look at the fbconfig(1) and related manpages (e.g. ffbconfig, m64config, pgxconfig, SUNWjfb_config, etc ...) for hardware framebuffer settings that may achieve the same effect. In general for non-Sun machines, look at the "-cc class" and related options in your X server manpage (perhaps Xserver(1)), it may allow modifying the default visual (e.g. "-cc 4", see for the visual class numbers). On XFree86 some video card drivers (e.g. Matrox mga) have settings like Option "Overlay" "24,8" to support multi-depth overlays. For these, use the "-cc 4" X server command line option to get a depth 24 default visual. Another option is if the system with overlay visuals is a Sun system running Solaris you can use the -overlay x11vnc option (Aug/2004) to have x11vnc use the Solaris XReadScreen(3X11) function to poll the "true view" of the whole screen at depth 24 TrueColor.This is useful for Legacy applications (older versions of Cadence CAD apps are mentioned by x11vnc users) that require the default depth be 8bpp, or will use a 8bpp visual even if depth 24 visuals are available, and so the default depth workaround described in the previous paragraph is not sufficient for these apps. The -overlay mode should also work on IRIX machines. Misc. notes on -overlay mode: An amusing by-product of -overlay mode is that mouse cursor shape is correct. The -overlay mode may be somewhat slower than normal mode due to the extra framebuffer manipulations that must be performed. Also, there is a bug in that for some popup menus, the windows they overlap will have painting problems while the popup is up (a workaround is to disable SaveUnders by passing -su to Xsun, e.g. in your /etc/dt/config/Xservers file). Still not working? Run xwininfo on the application with the messed up colors to verify that the depth of its visual is different from the default visual depth (gotten from xdpyinfo). One possible workaround in this case is to use the -id option to point x11vnc at the application window itself. If the application is complicated (lots of toplevel windows and popup menus) this may not be acceptable, and may even crash x11vnc (but not the application). It is theoretically possible to solve this problem in general (see xwd(1) for example), but it does not seem trivial or sufficiently fast for x11vnc to be able to do so in real time. Fortunately the -overlay option works for Solaris machines with overlay visuals where most of this problem occurs. Q-13: How do I figure out the window id to supply to the -id windowid option? Run the xwininfo program in a terminal. It will ask you to click on the desired application window. After clicking, it will print out much information, including the window id. Also, the visual and depth of the window printed out is often useful in debugging x11vnc [122]problems. When using -id windowid, note that some VNC viewers will have problems rendering screens that have a width that is not a multiple of 4. Try to manually adjust the window width before starting x11vnc -id .... Also, as of Dec/2004 libvncserver CVS you can use "-id pick" to have x11vnc run xwininfo(1) for you and after you click the window it extracts the windowid. Besides "pick" there is also "root" to allow you to go back to root window when doing remote-control. Q-14: Why don't menus or other transient windows come up when I am using the -id windowid option to view a single application window? This is related to the behavior of the XGetImage(3X11) and XShmGetImage() interfaces regarding backingstore, saveunders, etc. The way the image is retrieved depends on some aspects of how the X server maintains the display image data and whether other windows are clipping or obscuring it. See the XGetImage man page for more details. If you disable BackingStore and SaveUnders in the X server you should be able to see these transient windows. If things are not working and you still want to do the single window polling, try the -sid windowid option ("shifted" windowid). x11vnc is known to crash under both -id and -sid, so both modes are still experimental. Please report any reproducible bugs. Q-15: Can I use x11vnc to view and interact with an Xterminal (e.g. NCD) that is not running UNIX and so x11vnc cannot be run on it directly? You can, but it will likely be very wasteful of network bandwidth since you will be polling the X display over the network as opposed to over the local hardware. To do this, run x11vnc on a UNIX machine as close as possible network-wise to the Xterminal machine. Use the -display to point the display to that of the Xterminal (you'll of course need basic X11 permission to do that) and also supply the -noshm option (this enables the polling over the network). The response will likely be sluggish. This mode is not recommended except for "quick checks" of hard to get to X servers. Use something like -wait 150 to cut down on the polling rate. You may also need -flipbyteorder if the colors get messed up due to endian byte order differences. Q-16: Can I make x11vnc more quiet and also go into the background after starting up? Use the -q and -bg options, respectively. (also: -quiet is an alias for -q) Note that under -bg the stderr messages will be lost unless you use the "-o logfile" option. Q-17: How do I stop x11vnc once it is running in the background? If the -forever option has not been supplied, x11vnc will automatically exit after the first client disconnects. In general you will have to kill the x11vnc process (however, see the remote-control -R option below). This can be done via: "kill NNNNN" (where NNNNN is the x11vnc process id number found from ps(1)), or "pkill x11vnc", or "killall x11vnc" (Linux only, IIRC). If you have not put x11vnc in the background via the -bg option or shell & operator, then simply press Ctrl-C in the shell where x11vnc is running to stop it. If somehow your Keypress of Ctrl-C went through x11vnc to the Xserver that then delivered it to x11vnc it is possible one or both of the Ctrl or C keys will be left stuck in the pressed down state in the Xserver. Tapping the stuck key (either via x11vnc or at the physical console) will release it from the stuck state. If the keyboard seems to be acting strangely it is often fixed by tapping Ctrl, Shift, and Alt. Alternatively, the -clear_mods and -clear_keys options can be used to release pressed keys at startup and exit. As of Dec/2004 in the libvncserver CVS there is a remote control feature. It can change a huge amount of things on the fly: see the -remote and -query options. To shut down the running x11vnc server just type "x11vnc -R stop". To disconnect all clients do "x11vnc -R disconnect:all", etc. Q-18: Can I change settings in x11vnc without having to restart it? Is there a way to remote control it? Look at the -remote (same as -R) and -query (same as -Q) options added in the Dec/2004 libvncserver CVS. They allow nearly everything to be changed dynamically and settings to be queried. Examples: "x11vnc -R shared", "x11vnc -R forever", "x11vnc -R scale:3/4", "x11vnc -Q modtweak", "x11vnc -R stop", "x11vnc -R disconnect:all", etc.. These commands do not start a x11vnc server, but rather communicate with one that is already running. The X display (VNC_CONNECT property) is used as the communication channel, so the X permissions and DISPLAY must be set up correctly. There is also a simple tcl/tk gui based on this remote control mechanism. See the -gui option for more info. Q-19: Can I limit which machines incoming VNC clients can connect from? Yes, look at the -allow and -localhost options to limit connections by IP address. E.g. x11vnc -allow 192.168.0.1,192.168.0.2 for those two hosts or x11vnc -allow 192.168.0. for a subnet. For individual hosts you can use the hostname instead of the IP number. Note that -localhost is the same as "-allow 127.0.0.1" For more control, build libvncserver with libwrap support (tcp_wrappers) and then use /etc/hosts.allow. Q-20: How do I build x11vnc/libvncserver with libwrap (tcp_wrappers) support? Here is one way to pass this information to the configure script: env CPPFLAGS=-DUSE_LIBWRAP LDFLAGS=-lwrap ./configure then run make as usual. This requires libwrap and its development package (tcpd.h) to be installed on the build machine. If additional CPPFLAGS or LDFLAGS options are needed supply them as well. The resulting x11vnc then uses libwrap/tcp_wrappers for connections. The service name you will use in /etc/hosts.allow and /etc/hosts.deny is "vnc", e.g.: vnc: 192.168.100.3 .example.com Note that if you run x11vnc out of [123]inetd you do not need to build x11vnc with libwrap support because the /usr/sbin/tcpd reference in /etc/inetd.conf handles the tcp_wrappers stuff. Q-21: Can I prompt the user at the local X display whether the incoming VNC client should be accepted or not? Can I decide to make some clients view-only? How about running an arbitrary program to make the decisions? Yes, look at the "-accept command" option, it allows you to specify an external command that is run for each new client. (use quotes around the command if it contains spaces, etc.). If the external command returns 0 the client is accepted, otherwise the client is rejected. See below how to also accept clients view-only. The external command will have the RFB_CLIENT_IP environment variable set to the client's numerical IP address, RFB_CLIENT_PORT its port number. Similarly for RFB_SERVER_IP and RFB_SERVER_PORT to allow identification of the tcp virtual circuit. DISPLAY will be set to that of the X11 display being polled. Also, RFB_X11VNC_PID is set to the x11vnc process id (e.g. in case you decided to kill it), RFB_CLIENT_ID will be an id number, and RFB_CLIENT_COUNT the number of other clients currently connected. RFB_MODE will be "accept". As a special case, "-accept popup" will instruct x11vnc to create its own simple popup window. To accept the client press "y" or click mouse on the "Yes" button. To reject the client press "n" or click mouse on the "No" button. To accept the client View-only, press "v" or click mouse on the "View" button. If the -viewonly option has been supplied, the "View" action will not be present: the whole display is view only in that case. The popup window times out after 120 seconds, to change this behavior use "-accept popup:N" where N is the number of seconds (use 0 for no timeout). More tricks: "-accept popupmouse" will only take mouse click responses, while "-accept popupkey" will only take keystroke responses (popup takes both). After any of the 3 popup keywords you can supply a position of the window: +N+M, (the default is to center the window) e.g. -accept popupmouse+10+10. Also as a special case "-accept xmessage" will run the xmessage(1) program to prompt the user whether the client should be accepted or not. This requires that you have xmessage installed and available via PATH. In case it is not already on your system, the xmessage program is available at [124]ftp://ftp.x.org/ To include view-only decisions for the external commands, prefix the command something like this: "yes:0,no:*,view:3 mycommand ..." This associates the three actions: yes(accept), no(reject), and view(accept-view-only), with the numerical return codes. Use "*" instead of a number to set the default action (e.g. in case the external command returns an unexpected return code). Here is an example -accept script called accept_or_lock. It uses xmessage and xlock (replace with your screen lock command, maybe it is "xscreensaver-command -lock", or kdesktop_lock, or "dtaction LockDisplay"). It will prompt the user at the X display whether to accept, reject, or accept view-only the client, but if the prompt times out after 60 seconds the screen is locked and the VNC client is accepted. This allows the remote access when no one is at the display. #!/bin/sh # # accept_or_lock: prompt user at X display whether to accept an incoming # VNC connection. If timeout expires, screen is locked # and the VNC viewer is accepted (allows remote access # when no one is sitting at the display). # # usage: x11vnc ... -forever -accept 'yes:0,no:*,view:4 accept_or_lock' # xmessage -buttons yes:2,no:3,view-only:4 -center \ -timeout 60 "x11vnc: accept connection from $RFB_CLIENT_IP?" rc=$? if [ $rc = 0 ]; then xlock & sleep 5 exit 0 elif [ $rc = 2 ]; then exit 0 elif [ $rc = 4 ]; then exit 4 fi exit 1 Stefan Radman has written a nice dtksh script [125]dtVncPopup for use in CDE environments to do the same sort of thing. Information on how to use it is found at the top of the file. He encourages you to provide feedback to him to help improve the script. Note that in all cases x11vnc will block while the external command or popup is being run, so attached clients will not receive screen updates, etc during this period. (use of -threads may or may not alter this behavior). To run a command when a client disconnects, use the "-gone command" option. This is for the user's convenience only: the return code of the command is not interpreted by x11vnc. The same environment variables are set as in "-accept command" (except that RFB_MODE will be "gone"). Q-22: How do I create a VNC password for use with x11vnc? You may already have one in $HOME/.vnc/passwd if you have used, say, the vncserver program from the regular RealVNC or TightVNC packages (i.e. launching the Xvnc server). Otherwise, you could use the vncpasswd(1) program from those packages. The libvncserver package also comes with a simple program: storepasswd in the examples directory. And as of Jun/2004 in the libvncserver CVS x11vnc supports the -storepasswd "pass" "file" option, which is the the same functionality of storepasswd. Be sure to quote the "pass" if it contains shell meta characters, spaces, etc. You then use the password via the x11vnc option: -rfbauth filename Compared to vncpasswd(1) the latter two methods are a somewhat unsafe because the password is specified on the command line and so someone may see it by using ps(1) or looking over your shoulder. Also watch out for the command winding up in your shell's history file (history -c is often a way to clear it). x11vnc also has the [126]-passwdfile and -passwd/-viewpasswd plain text (i.e. not obscured like the -rfbauth VNC passwords) password options. Q-23: How can I tunnel my connection to x11vnc via an encrypted SSH channel between two Unix machines? See the description earlier on this page on [127]how to tunnel VNC via SSH from Unix to Unix. A number of ways are described along with some issues you may encounter. Other secure encrypted methods exists, e.g. stunnel, IPSEC, various VPNs, etc. Q-24: How can I tunnel my connection to x11vnc via an encrypted SSH channel from Windows using an SSH client like Putty? [128]Above we described how to tunnel VNC via SSH from Unix to Unix, you may want to review it. To do this from Windows using Putty it would go something like this: * In the Putty dialog window under 'Session' enter the hostname or IP number of the Unix machine with display to be viewed. * Make sure the SSH protocol is selected and the server port is correct. * Under 'Connections/SSH/Tunnels' Add a Local connection with 'Source port: 5900' and 'Destination: localhost:5900' * Log into the remote machine by pressing 'Open' and supplying username, password, etc. * In that SSH shell, start up x11vnc by typing the command: x11vnc -display :0 plus any other desired options. * Finally, start up your VNC Viewer in Windows and enter 'localhost:0' as the VNC server. You can keep all of the settings in a Putty 'Saved Session'. Also, once everything is working, you can consider putting x11vnc -display :0 (plus other cmdline options) in the 'Remote command' Putty setting under 'Connections/SSH'. It is likely possible to script the whole process in a BAT file including launching the VNC viewer by using the plink Putty utility. Send us the script if you get that working. For extra protection feel free to run x11vnc with the -localhost and -rfbauth/-passwdfile options. If the machine you SSH into via Putty is not the same machine with the X display you wish to view (e.g. your company provides incoming SSH access to a gateway machine), then you need to change the above Putty dialog setting to: 'Destination: otherhost:5900', Once logged in, you'll need to do a second login (ssh or rsh) to the workstation machine 'otherhost' and then start up x11vnc on it. As discussed [129]above another option is to first start the VNC viewer in "listen" mode, and then launch x11vnc with the "-connection localhost" option to establish the reverse connection. In this case a Remote port redirection (not Local) is needed for port 5500 instead of 5900 (i.e. 'Source port: 5500' and 'Destination: localhost:5500' for a Remote connection). Q-25: Does x11vnc support Unix usernames and passwords? Can I further limit the set of Unix usernames who can connect to the VNC desktop? Until the VNC protocol and libvncserver support this things will be approximate at best. Hopefully, it will not be too long to wait for such support. One approximate method involves starting x11vnc with the -localhost option. This basically requires the viewer user to log into the workstation where x11vnc is running via their Unix username and password, and then somehow set up a port redirection of his vncviewer connection to make it appear to emanate from the local machine. As discussed above, ssh is useful for this: ssh -l username -L 5900:localhost:5900 hostname ... See the ssh wrapper scripts mentioned [130]elsewhere on this page. Of course a malicious user could allow other users to get in through his channel, but that is a problem with every method. Another thing to watch out for is a malicious user on the viewer side (where ssh is running) trying to sneak in through the ssh port redirection. Regarding limiting the set of Unix usernames who can connect, the traditional way would be to further require a VNC password to supplied (-rfbauth, -passwd, etc). A scheme that avoids a second password involves using the -accept option that runs a program to examine the connection information to determine which user is connecting from the local machine. For example, the program could use the ident service on the local machine (normally ident should not be trusted over the network, but on the local machine it should be accurate: otherwise root has been compromised and so there are more serious problems!). An example script passed in via -accept scriptname that deduces the Unix username and limits who can be accepted might look something like this: #!/bin/sh if [ "$RFB_CLIENT_IP" != "127.0.0.1" -o "$RFB_SERVER_IP" != "127.0.0.1" ]; then exit 1 # something fishy... reject it. fi user=`echo "$RFB_CLIENT_PORT, $RFB_SERVER_PORT" | nc -w 1 $RFB_CLIENT_IP 113 \ | grep 'USERID.*UNIX' | head -1 | sed -e 's/[\r ]//g' | awk -F: '{print $4}'` for okuser in fred barney wilma betty do if [ "X$user" = "X$okuser" ]; then exit 0 # accept it fi done exit 1 # reject it For this to work with ssh port redirection, the ssh option UsePrivilegeSeparation must be enabled. Q-26: Can I have two passwords for VNC viewers, one for full access and the other for view-only access to the display? Yes, as of May/2004 in the libvncserver CVS there is the -viewpasswd option to supply the view-only password. Note the full-access password option -passwd must be supplied at the same time. E.g.: -passwd sword -viewpasswd fish. To avoid specifying the passwords on the command line (where they could be observed via the ps(1) command by any user) you can use the -passwdfile option to specify a file containing plain text passwords. Presumably this file is readable only by you, and ideally it is located on the machine x11vnc is run on (to avoid being snooped on over the network). The first line of this file is the full-access password. If there is a second line in the file and it is non-blank, it is taken as the view-only password. View-only passwords currently do not work for the -rfbauth password option (standard VNC password storing mechanism). FWIW, note that although the output (usually placed in $HOME/.vnc/passwd) by the vncpasswd or storepasswd programs (or from x11vnc -storepasswd) looks encrypted they are really just obscured to avoid "casual" password stealing. It takes almost no skill to figure out how to extract the plain text passwords from $HOME/.vnc/passwd since it is very straight-forward to work out what to do from the VNC source code. Q-27: I use a screen-lock when I leave my workstation (e.g. xscreensaver or xlock). When I remotely access my workstation desktop via x11vnc I can unlock the desktop fine, but I am worried people will see my activities on the physical monitor. What can I do to prevent this, or at least make it more difficult? Probably most work environments would respect your privacy if you powered off the monitor. Also remember if people have physical access to your workstation they basically can do anything they want with it (e.g. install a backdoor for later use, etc). In any event, as of Jun/2004 there is an experimental utility to make it more difficult for nosey people to see your x11vnc activities. The source for it is [131]blockdpy.c The idea behind it is simple (but obviously not bulletproof): when a VNC client attaches to x11vnc put the display monitor in the DPMS "off" state, if the DPMS state ever changes immediately start up the screen-lock program. The x11vnc user will notice something is happening and think about what to do next (while the screen is in a locked state). This works (or at least has a chance of working) because if the intruder moves the mouse or presses a key on the keyboard, the monitor wakes up out of the DPMS off state, and this induces the screen lock program to activate as soon as possible. Of course there are cracks in this, the eavesdropper could detach your monitor and insert a non-DPMS one, and there are race conditions. As mentioned above this is not bulletproof. A really robust solution would likely require X server and perhaps even video hardware support. The blockdpy utility is launched by the -accept option and told to exit via the -gone option (the vnc client user should obviously re-lock the screen before disconnecting!). Instructions can be found in the source code for the utility at the above link. Q-28: Can I have x11vnc automatically lock the screen when I disconnect the VNC viewer? Yes, a user mentions he uses the -gone option under CDE to run a screen lock program: x11vnc -display :0.0 -forever -gone 'dtaction LockDisplay' Other possibilities are: x11vnc -display :0.0 -forever -gone 'xscreensaver-command -lock' x11vnc -display :0.0 -forever -gone 'kdesktop_lock' x11vnc -display :0.0 -forever -gone 'xlock &' Q-29: Are reverse connections (i.e. the VNC server connecting to the VNC viewer) using "vncviewer -listen" and vncconnect(1) supported? As of Mar/2004 in the libvncserver CVS x11vnc supports reverse connections. On Unix one starts the VNC viewer in listen mode: vncviewer -listen, and then startus up x11vnc with the -connect option. To connect immediately at x11vnc startup time use the "-connect host:port" option (use commas for a list of hosts to connect to). The ":port" is optional (default is 5500). If a file is specified instead: -connect /path/to/some/file then that file is checked periodically (about once a second) for new hosts to connect to. To use the vncconnect(1) program (from the core VNC package at www.realvnc.com) specify the -vncconnect option to x11vnc (Note: as of Dec/2004 -vncconnect is now the default). vncconnect(1) must be pointed to the same X11 DISPLAY as x11vnc (since it uses X properties to communicate with x11vnc). If you do not have or do not want to get the vncconnect(1) program, the following script (named "Vncconnect") may work if your xprop(1) supports the -set option: #!/bin/sh # usage: Vncconnect # Vncconnect # note: not all xprop(1) support -set. # xprop -root -f VNC_CONNECT 8s -set VNC_CONNECT "$1" Q-30: Sometimes when a VNC viewer dies abruptly, x11vnc also dies with the error message like: "Broken pipe". I'm using the -forever mode and I want x11vnc to keep running. As of Jan/2004 in the libvncserver CVS the SIGPIPE signal is ignored. So if a viewer client terminates abruptly, libvncserver will notice on the next I/O operation and will close the connection and continue on. As of Apr/2004 the above fix only works for BSD signal systems (Linux, FreeBSD, ...) For SYSV systems there is a workaround in my [132]x11vnc.c file. It also has an option -sigpipe exit to have x11vnc clean up and exit upon receiving SIGPIPE. Q-31: How can I use x11vnc to connect to an X login screen like xdm, GNOME gdm, KDE kdm, or CDE dtlogin? (i.e. nobody is logged into an X session yet). One time only. If the X login screen is running and you just want to connect to it once: On Linux it seems it is possible to do this by just adjusting the XAUTHORITY to point to the MIT-COOKIE auth file while running x11vnc as root, e.g. for the gnome display manager, gdm: env XAUTHORITY=/var/gdm/:0.Xauth x11vnc -display :0 or equivalently with newer x11vnc: x11vnc -auth /var/gdm/:0.Xauth -display :0 There will be a similar thing for xdm using however a different auth directory path (perhaps something like /var/lib/xdm/authdir/authfiles/A:0-XQvaJk for xdm or /var/lib/kdm/A:0-crWk72 for kdm, where the random characters in basename will vary a bit). Another place to look for the auth file is via ps(1) output for the X server process to find any -auth argument. gdm seems to have an annoying setting that causes x11vnc (and any other X clients) to be killed after the user logs in. Setting KillInitClients=false in /etc/X11/gdm/gdm.conf may avoid this. For dtlogin in addition to the above sort of trick (the auth file should be in /var/dt), you'll also need to add something like Dtlogin*grabServer:False to the Xconfig file (/etc/dt/config/Xconfig or /usr/dt/config/Xconfig on Solaris). Then restart dtlogin, e.g.: /etc/init.d/dtlogin stop; /etc/init.d/dtlogin start or reboot. Continuously. Have x11vnc reattach each time the X server is restarted (i.e. after each logout): To make x11vnc always attached to the the X server (please consider the security implications of this!) including the login screen you will need to add a command to the display manager startup script. The name of the script file depends on desktop used and seem to be: GNOME /etc/X11/gdm/PreSession/Default KDE /etc/kde*/kdm/Xsetup XDM /etc/X11/xdm/Xsetup CDE /etc/dt/config/Xsetup although the exact location can depend on operating system and distribution. See the documentation for your display manager: gdm(1), kdm(1), xdm(1), dtlogin(1) for additional details. The above Dtlogin*grabServer:False step will be needed for dtlogin here as well. There may also be display number specific scripts: e.g. Xsetup_0 vs. Xsetup. In any event, the line you will add to the script will look something like: /usr/local/bin/x11vnc -rfbauth /path/to/my/vnc/passwd -forever -bg where you should customize the exact command to your needs. You may want to redirect the x11vnc output to a separate log file for debugging. In that case replace the -bg above with something like: 1>> /var/tmp/x11vnc.log 2>&1 & (or use the "-o logfile" option). If you do not want to deal with the display manager startup scripts, here is a kludgey script that can be run manually or out of a boot file like rc.local. [133]x11vnc_loop It will need some local customization before running. Because the XAUTHORITY auth file must be guessed by this script, use of the display manager script above is preferred. Xterminals: Note that if the display machine is a traditional Xterminal (where the X server runs on the Xterminal box, but all of the X client applications run on a central server), the login display manager, and hence the auth files, are on the central server and not on the Xterminal where the X server and x11vnc processes are running. Somehow the MIT-COOKIE auth file data must be copied to the Xterminal (e.g. via ssh, possibly using xauth nextract). You can use xauth -f cookie.file list to test the contents of the cookie in a file "cookie.file" Note: use of Display Manager (gdm, kdm, ...) auth cookie files (i.e. from /var/..., /tmp/..., or elsewhere) may require modification via xauth(1) to correctly include the display x11vnc refers to (e.g. xauth -f cookie.file add :0 . 45be51ae2ce9dfbacd882ab3ef8e96b1, where 45be51... was found from an xauth -f /var/... list) or other reasons. See xauth(1) manpage for full details on how to transfer an MIT-MAGIC-COOKIE between machines and displays. Q-32: Can I run x11vnc out of inetd(1)? Yes, perhaps a line something like this in /etc/inetd.conf will do it for you: 5900 stream tcp nowait root /usr/sbin/tcpd /usr/local/bin/x11vnc_sh where the shell script /usr/local/bin/x11vnc_sh looks something like (you'll need to customize to your settings). #!/bin/sh XAUTHORITY=/home/fred/.Xauthority; export XAUTHORITY /usr/local/bin/x11vnc -inetd -display :0 -rfbauth /home/fred/.vnc/passwd 2>> /t mp/x11vnc_sh.log Note that you must redirect the standard error output to a log file or /dev/null for proper operation via inetd (otherwise the standard error also goes to the VNC vncviewer, and that confuses it greatly). If you do not use a wrapper script as above but rather call x11vnc directly in /etc/inetd.conf and do not redirect stderr, then you must specify the -q (aka -quiet) option: /usr/local/bin/x11vnc -q -inetd ... or use the -o logfile to collect the output in a file. The wrapper script with stderr redirection to a log file is the recommended method because the errors and warnings printed out are very useful in troubleshooting problems. Note also the need to set XAUTHORITY to point to the MIT-COOKIE auth file to get permission to connect to the X display (the x11vnc -auth option could also have been used). See the x11vnc_loop file in the previous question for more ideas on what that auth file may be, etc. On Solaris you cannot have the bare number 5900 in /etc/inetd.conf, you'll need to replace it with a word like x11vnc an then put something like "x11vnc 5900/tcp" in /etc/services. Be sure to look at your /etc/hosts.allow and /etc/hosts.deny settings to limit the machines that can connect to this service (your desktop!). For the above example with /etc/hosts.allow: x11vnc_sh : 123.45.67.89 A really safe way to do things is to limit the above inetd to localhost only (via /etc/hosts.allow) and use ssh to tunnel the incoming connection. Using inetd for this prevents there being a tiny window of opportunity between x11vnc starting up and your vncviewer connecting to it. Always use a VNC password to further protect against unwanted access. Q-33: How do I make x11vnc work with the Java VNC viewer applet in a web browser? To have x11vnc serve up a Java VNC viewer applet to any web browsers that connect to it, run x11vnc with this option: -httpdir /path/to/the/java/classes/dir (this directory will contain the files index.vnc and, for example, VncViewer.jar) Note that libvncserver contains the TightVNC Java classes jar file for your convenience. (it is the file classes/VncViewer.jar in the source tree). You will see output something like this: 14/05/2004 11:13:56 Autoprobing selected port 5900 14/05/2004 11:13:56 Listening for HTTP connections on TCP port 5800 14/05/2004 11:13:56 URL http://walnut:5800 14/05/2004 11:13:56 screen setup finished. 14/05/2004 11:13:56 The VNC desktop is walnut:0 PORT=5900 then you can connect to that URL with any Java enabled browser. Feel free to customize the default index.vnc file in the classes directory. Note that if you wanted to, you could also start the Java viewer entirely from the viewer-side by having the jar file there and using either the java or appletviewer commands to run the program. Q-34: Why isn't the mouse cursor shape (the little icon shape where the mouse pointer is) correct as I move from window to window? As mentioned above, the X11 mouse cursor shape (i.e. little picture: an arrow, X, I-beam, resizer, etc) is one of the few write-only objects in X11. That is, an application can tell the X server what the cursor shape should be when the pointer is in a given window, but a program (like x11vnc) unfortunately cannot read this information. I believe this is because the cursor shape is often downloaded to the graphics hardware (video card), but I could be mistaken. A simple kludge is provided by the "-cursor X" option that changes the cursor when the mouse is on the root background (or any window has the same cursor as the root background). Note that desktops like GNOME or KDE often cover up the root background, so this won't work for those cases. Also see the "-cursor some" option for additional kludges. It should be possible to apply some heuristics where x11vnc tries to build up a table of cursors for the windows it sees, perhaps using a user supplied hints file, but that has not been explored yet. Also note that as of Aug/2004 in the libvncserver CVS, on Solaris using the SUN_OVL overlay extension, x11vnc can show the correct mouse cursor when the -overlay is also supplied. (-overlay has some other problems however, and can be slower). Also as of Dec/2004 in libvncserver CVS XFIXES X extension support has been added to allow exact extraction of the mouse cursor shape. This is the preferred way of viewing the cursor shape, it is available on recent Linux Xorg based distros and Solaris 10 express (on Solaris you will need to add "-L /usr/openwin/sfw/lib -R /usr/openwin/sfw/lib" to LDFLAGS for configure to enable it). Q-35: Why does the mouse arrow just stay in one corner in my vncviewer, whereas my cursor (that does move) is just a dot? This default takes advantage of a [134]tightvnc extension that allows specifying a cursor image for the local VNC viewer. You may disable it with the -nocursor option to x11vnc if your viewer does not have this extension. Note: as of Aug/2004 in the libvncserver CVS this should be fixed: the default for non-tightvnc viewers will be to draw the moving cursor into the framebuffer. This can also be disabled via -nocursor. Q-36: Can I take advantage of the TightVNC extension to the VNC protocol where Cursor Positions Updates are sent back to all connected clients (i.e. passive viewers can see the mouse cursor being moved around by another viewer)? Use the -cursorpos option when starting x11vnc. A VNC viewer must support the Cursor Positions Updates for the user to see the mouse motions (the TightVNC viewers support this). As of Aug/2004 in the libvncserver CVS -cursorpos is the default. See -nocursorpos and -nocursorshape. Q-37: Is it possible to swap the mouse buttons (e.g. left-handed operation), or arbitrarily remap them? How about mapping button clicks to keystrokes, e.g. to partially emulate Mouse wheel scrolling? You can remap the mouse buttons via something like: -buttonmap 13-31 (or perhaps 12-21). Also, note that xmodmap(1) lets you directly adjust the X server's button mappings, but in some circumstances it might be more desirable to have x11vnc do it. One user had an X server with only one mouse button(!) and was able to map all of the VNC client mouse buttons to it via: -buttonmap 123-111. Note that the -debug_pointer option prints out much info for every mouse/pointer event. To map mouse button clicks to keystrokes you can use the alternate format where the keystrokes are enclosed between colons like this :: in place of the mouse button digit. For a sequence of keysyms separate them with "+" signs. Button clicks can also be included in the sequence via the fake keysyms Button1, etc. As an example, suppose the VNC viewer machine has a mouse wheel (these generate button 4 and 5 events), but the machine that x11vnc is run on only has the 3 regular buttons. In normal operation x11vnc will discard the button 4 and 5 events. However, either of the following button maps could possibly be of use emulating the mouse wheel events in this case: -buttonmap 12345-123:Prior::Next: -buttonmap 12345-123:Up+Up+Up::Down+Down+Down: Exactly what keystroke "scrolling" events they should be bound to depends on one's taste. If this method is too approximate, one could consider not using -buttonmap but rather configuring the X server to think it has a mouse with 5 buttons even though the physical mouse does not. Note that when a keysym-mapped mouse button is clicked down this immediately generates the key-press and key-release events (for each keysym in turn if the mapping has a sequence of keysyms). When the mouse button goes back up nothing is generated. If you include modifier keys like Shift_L instead of key-press immediately followed by key-release the state of the modifier key is toggled (however the initial state of the modifier key is ignored). So to map the right button to type my name 'Karl Runge' I could use this: -buttonmap 3-:Shift_L+k+Shift_L+a+r+l+space+Shift_L+r+Shift_L+u+n+g+e: (yes, this is getting a little silly). Q-38: When I drag windows around with the mouse or scroll up and down things really bog down (unless I do the drag in a single, quick motion). Is there anything to do to improve things? As of Jan/2004 there are some improvements in the libvncserver CVS tree. The default should now be much better than before and dragging small windows around should no longer be a huge pain. If for some reason these changes make matters worse, you can go back to the old way via the -old_pointer option. Also added was the -nodragging option that disables all screen updates while dragging with the mouse (i.e. mouse motion with a button held down). This gives the snappiest response, but might be undesired in some circumstances when you want to see the visual feedback while dragging (e.g. menu traversal or text selection). As of Dec/2004 in the libvncserver CVS the -pointer_mode n option was introduced. n=1 is the -old_pointer. See the -pointer_mode n help for more info. Q-39: I have lots of memory, but why does x11vnc fail with shmget: No space left on device or Minor opcode of failed request: 1 (X_ShmAttach)? It is not a matter of free memory, but rather free shared memory (shm) slots, also known as shm segments. This often occurs on a public Solaris machine using the default of only 100 slots. You (or the owner or root) can clean them out with ipcrm(1). x11vnc tries hard to release its slots, but it, and other programs, are not always able to (e.g. if kill -9'd). Sometimes x11vnc will notice the problem with shm segments and tries to get by with fewer, only giving a warning like this: 19/03/2004 10:10:58 shmat(tile_row) failed. shmat: Too many open files 19/03/2004 10:10:58 error creating tile-row shm for len=4 19/03/2004 10:10:58 reverting to single_copytile mode Here is a shell script [135]shm_clear to list and prompt for removal of your unattached shm segments (attached ones are skipped). I use it while debugging x11vnc (I use shm_clear -y to assume "yes" for each prompt). If x11vnc is regularly not cleaning up its shm segments, please contact me so we can work to improve the situation. Longer term, on Solaris you can put something like this in /etc/system: set shmsys:shminfo_shmmax = 0x2000000 set shmsys:shminfo_shmmni = 0x1000 to sweep the problem under the rug (4096 slots). On Linux, examine /proc/sys/kernel/shmmni; you can modify the value by writing to that file. Things are even more tight on Solaris 8 and earlier, there is a default maximum number of shm segments per process of 6. The error is the X server (not x11vnc) being unable to attach to the segments, and looks something like this: 30/04/2004 14:04:26 Got connection from client 192.168.1.23 30/04/2004 14:04:26 other clients: X Error of failed request: BadAccess (attempt to access private resource den ied) Major opcode of failed request: 131 (MIT-SHM) Minor opcode of failed request: 1 (X_ShmAttach) Serial number of failed request: 14 Current serial number in output stream: 17 This tight limit on Solaris 8 can be increased via: set shmsys:shminfo_shmseg = 100 in /etc/system. See the next paragraph for more workarounds. To minimize the number of shm segments used by x11vnc try using the -onetile option (corresponds to only 3 shm segments used, and adding -fs 1.0 knocks it down to 2). If you are having much trouble with shm segments, consider disabling shm completely via the -noshm option. Performance will be somewhat degraded but when done over local machine sockets it should be acceptable (see an [136]earlier question discussing -noshm). Q-40: How can I make x11vnc use less system resources? The -nap and "-wait n" (where n is the sleep between polls in milliseconds, the default is 30 or so) option are good places to start. Reducing the X server bits per pixel depth (e.g. to 16bpp or even 8bpp) will further decrease memory I/O and network I/O. Using the -onetile option will use less memory and use fewer shared memory slots (add -fs 1.0 for one less slot). Q-41: How can I make x11vnc use MORE system resources? You can try -threads and dial down the wait time (e.g. -wait 1) and possibly dial down -defer as well. Note that if you try to increase the "frame rate" too much you can bog down the server end with the extra work it needs to do compressing the framebuffer data, etc. That said, it is possible to "stream" video via x11vnc if the video window is small enough. E.g. a 256x192 xawtv TV capture window (using the x11vnc -id option) can be streamed over a LAN or wireless at a reasonable frame rate. Q-42: I use x11vnc over a slow link with high latency (e.g. dialup modem), is there anything I can do to speed things up? Some things you might want to experiment with (most of which will help performance on faster links as well): X server/session parameters: * Configure the X server bits per pixel to 16bpp or even 8bpp. (reduces amount of data needed to be sent) * Use a smaller desktop size (e.g. 1024x768 instead of 1280x1024) * Make sure the desktop background is a solid color (the background is resent every time it is re-exposed) * Configure your window manager or desktop "theme" to not use fancy images, shading, and gradients for the window decorations, etc. Disable Opaque moves, resizes, and animations. * Avoid small scrolls of large windows using the Arrow keys or scrollbar. Try to use PageUp/PageDown instead. * On XFree86 turn on the Shadow Framebuffer to speed up reading. (Option "ShadowFB" "true", and possibly Option "NoAccel" "true" as well, in the Device section of /etc/X11/XF86Config) This disables 2D acceleration on the physical display and so is probably not worth it, but could be of use in some situations. VNC viewer parameters: * Use a [137]TightVNC enabled viewer! * Make sure the tight encoding is being used (look at vncviewer and x11vnc outputs) * Request 8 bits per pixel using -bgr233 (up to 4X speedup over depth 24 TrueColor (32bpp), but colors will be off) * The -depth option is similar to -bgr233 in reducing the requested number of bits per pixel. * Try increasing the TightVNC -compresslevel (compresses more on server side before sending, but uses more CPU) * Try reducing the TightVNC -quality (increases JPEG compression, but is lossy) * Try other VNC encodings via -encodings (tight is probably fastest, but you should check) * On the machine where vncviewer is run, make sure Backing Store is enabled (XFree86 disables it by default causing re-exposures of vncviewer to be very slow) x11vnc parameters: * Try using -nodragging (no screen updates when dragging mouse, but sometimes you miss visual feedback) * Try the -progressive pixelheight mode with the block pixelheight 100 or so (delays sending vertical blocks since they may change while viewer is receiving earlier ones) * Set -fs 1.0 (disables fullscreen updates) * Enable server-side scaling via -scale m/n to make the VNC frambuffer smaller. (less data needs to be sent, but image may not be sharp and requires extra resources on the server end, may actually slow down transfers of text due to poorer compression) * Try increasing -wait or -defer (reduces the maximum "frame rate", but won't help much for large screen changes) * If you just want to watch one (simple) window use -id (cuts down extraneous polling and updates, but can be buggy or insufficient) * Set -nosel (disables all clipboard selection exchange) * Use -nocursor (repainting the remote cursor position and shape takes resources and round trips) Q-43: How can I get my AltGr and Shift modifiers to work between keyboards for different languages? The option -modtweak should be of some use for this. It is a mode that monitors the state of the Shift and AltGr Modifiers and tries to deduce the correct keycode to send, possibly by sending fake modifier key presses and releases in addition to the actual keystroke. Update: As of Jul/2004 in the libvncserver CVS, -modtweak is now the default (use -nomodtweak to get the old behavior). This was done because it was noticed on newer XFree86 setups even on bland "us" keyboards like "pc104 us" XFree86 included a "ghost" key with both "<" and ">" it. This key does not exist on the keyboard (see [138]this FAQ for more info). Without -modtweak there was then an ambiguity in the reverse map keysym => keycode, making it so the "<" symbol could not be typed. Also see the [139]FAQ about the -xkb option for a more powerful method of modifier tweaking for use on X servers with the XKEYBOARD extension. When trying to resolve keyboard mapping problems, that the -debug_keyboard option prints out much info for every keystroke and so can be useful debugging things. Q-44: When I try to type a "<" (i.e. less than) instead I get ">" (i.e. greater than)! Strangely, typing ">" works OK!! Does your keyboard have a single key with both "<" and ">" on it? Even if it doesn't, your X server may think your keyboard has such a key (e.g. pc105 in the XF86Config file when it should be something else, say pc104). The problem here is that on the Xserver where x11vnc is run there are two keycodes that correspond to the "<" keysym. Run something like this to see: xmodmap -pk | egrep -i 'KeyCode|less|greater' There are 4 KeySyms per KeyCode; KeyCodes range from 8 to 255. KeyCode Keysym (Keysym) ... 59 0x002c (comma) 0x003c (less) 60 0x002e (period) 0x003e (greater) 94 0x003c (less) 0x003e (greater) That keycode 94 is the special key with both "<" and ">". When x11vnc receives the "<" keysym over the wire from the remote VNC client, it unfortunately maps it to keycode 94 instead of 59, and sends 94 to the X server. Since Shift is down (i.e. you are Shifting the comma key), the X server interprets this as Shifted-94, which is ">". A workaround in the X server configuration is to "deaden" that special key: xmodmap -e "keycode 94 = " However, one user said he had to do this: xmodmap -e "keycode 94 = 0x002c 0x003c" (If the numerical values are different for your setup, substitute the ones that correspond to your display. The above xmodmap scheme can often be used to work around other ambiguous keysym to keycode mappings). Alternatively, here are some x11vnc options to try to work around the problem: -modtweak and -remap less-comma These are convenient in that they do not modify the actual X server settings. The former is a mode that monitors the state of the Shift and AltGr modifiers and tries to deduce the correct keycode sequence to send. The latter is an immediate remapping of the keysym less to the keysym comma when it comes in from a client (so when Shift is down the comma press will yield "<"). See also the [140]FAQ about the -xkb option as a possible workaround using the XKEYBOARD extension. Note that of Jul/2004 in the libvncserver CVS the -modtweak option is now that default. Note that the -debug_keyboard option prints out much info for every keystroke to aid debugging keyboard problems. Q-45: I'm using an "international" keyboard (e.g. German "de", or Danish "dk") and the -modtweak mode works well if the VNC viewer is run on a Unix/Linux machine with a similar keyboard. But if I run the VNC viewer on Unix/Linux with a different keyboard (e.g. "us") or Windows with any keyboard, I can't type some keys like: "@", "$", "<", ">", etc. How can I fix this? The problem with Windows is it does not seem to handle AltGr well. It seems to fake it up by sending Control_L+Alt_R to applications. The Windows VNC viewer sends those two down keystrokes out on the wire to the VNC server, but when the user types the next key to get, e.g., "@" the Windows VNC viewer sends events bringing the up the Control_L+Alt_R keys, and then sends the "@" keysym by itself. The Unix/Linux VNC viewer on a "us" keyboard does a similar thing since "@" is the Shift of the "2" key. The keysyms Shift and "@" are sent to the VNC server. In both cases no AltGr is sent to the VNC server, but we know AltGr is needed on the physical international keyboard to type a "@". This all worked fine with x11vnc running with the -modtweak option (it figures out how to adjust the Modifier keys (Shift or AltGr) to get the "@"). However it fails under recent versions of XFree86 (and the X.org fork). These run the XKEYBOARD extension by default and make heavy use of it to handle international keyboards. To make a long story short, on these newer XFree86 setups the traditional X keymap lookup x11vnc uses is no longer accurate. x11vnc can't find the keysym "@" anywhere in the keymapping! (even though it is in the XKEYBOARD extended keymapping). How to Solve: As of Jul/2004 in the libvncserver CVS x11vnc has two changes: * -modtweak (tweak Modifier keys) is now the default (use -nomodtweak to go back to the old way) * there is a new option -xkb to use the XKEYBOARD extension API to do the Modifier key tweaking. The -xkb option seems to fix all of the missing keys: "@", "<", ">", etc.: it is recommended that you try it if you have this sort of problem. Let us know if there are any remaining problems (see the next paragraph for some known problems). If you specify the -debug_keyboard (aka -dk) option twice you will get a huge amount of keystroke debugging output (send it along with any problems you report). Known problems: * One user had to disable a "ghost" Mode_switch key that was causing problems under -xkb. His physical AltGr key was bound to ISO_Level3_Shift (which seems to be the XKEYBOARD way of doing things), while there was a ghost key Mode_switch (which seems to be obsolete) in the mapping as well. Both of these keysyms were bound to Mod5 and x11vnc was unfortunately choosing Mode_switch. From the x11vnc -xkb -dk -dk output it was noted that Mode_switch was attached to keycode 93 (no physical key generates this keycode) while ISO_Level3_Shift was attached to keycode 113. The keycode skipping option was used to disable the ghost key: -skip_keycodes 93 * In implementing -xkb we noticed that some characters were still not getting through, e.g. "~" and "^". This is not really an XKEYBOARD problem. What was happening was the VNC viewer was sending the keysyms asciitilde and asciicircum to x11vnc, but on the X server with the international keyboard those keysyms were not mapped to any keys. So x11vnc had to skip them. The way these characters are typically entered on international keyboards is by "dead" (aka "mute") keys. E.g. to enter "~" at the physical display the keysym dead_tilde is pressed and released (this usually involves holding AltGr down while another key is pressed) and then space is pressed. (this can also be used get characters with the "~" symbol on top, e.g. "ã" by typing "a" instead of space). What to do? In general the VNC protocol has not really solved this problem: what should be done if the VNC viewer sends a keysym not recognized by the VNC server side? Workarounds can possibly be created using the -remap x11vnc option: -remap asciitilde-dead_tilde,asciicircum-dead_circumflex etc. Use -remap filename if the list is long. Please send us your workarounds for this problem on your keyboard. Perhaps we can have x11vnc adjust automatically at some point. * To complement the above workaround using the -remap, an option -add_keysyms was added. This option instructs x11vnc to bind any unknown Keysyms coming in from VNC viewers to unused Keycodes in the X server. This modifies the global state of the X server. When x11vnc exits it removes the extra keymappings it created. Note that the -remap mappings are applied first, right when the Keysym is received from a VNC viewer, and only after that would -add_keysyms, or anything else, come into play. Q-46: When typing I sometimes get double, triple, or more of my keystrokes repeated. I'm sure I only typed them once, what can I do? This may be due to an interplay between your X server's key autorepeat delay and the extra time delays caused by x11vnc processing. Short answer: disable key autorepeating by running the command "xset r off" on the Xserver where x11vnc is run (restore via "xset r on") or use the new (Jul/2004) -norepeat x11vnc option. You will still have autorepeating because that is taken care of on your VNC viewer side. Update: as of Dec/2004 -norepeat is now the default. Use -repeat to disable it. Details: suppose you press a key DOWN and it generates changes in large regions of the screen. The CPU and I/O work x11vnc does for the large screen change could be longer than your X server's key autorepeat delay. x11vnc may not get to processing the key UP event until after the screen work is completed. The X server believes the key has been held down all this time, and applies its autorepeat rules. Even without inducing changes in large regions of the screen, this problem could arise when accessing x11vnc via a dialup modem or otherwise high latency link (e.g. > 250 ms latency). Look at the output of "xset q" for the "auto repeat delay" setting. Is it low (e.g. < 500 ms)? If you increase it to a very large delay, e.g. "xset r rate 1500" or turn off autorepeat completely: "xset r off", does the problem go away? We are looking into a good way to fix this, but in the meantime as a workaround try to use xset r rate N for large enough delay N or xset r off to get rid of the problem. Also, as of Jul/2004 in the libvncserver CVS there is the -norepeat option to automatically disable autorepeating in the X server whenever VNC viewers are attached. Note that the VNC viewer side of the connection will (nearly always) do its own autorepeating so there is no big loss here, unless someone is also working at the physical display and misses his autorepeating. Q-47: The machine where I run x11vnc has an AltGr key, but the local machine where I run the VNC viewer does not. Is there a way I can map a local unused key to send an AltGr? How about a Compose key as well? Something like "-remap Super_R-Mode_switch" x11vnc option may work. Note that Super_R is the "Right Windoze(tm) Flaggie" key; you may want to choose another. The -debug_keyboard option comes in handy in finding keysym names (so does xev(1)). For Compose how about "-remap Menu-Multi_key" (note that Multi_key is the official name for Compose) To do both at the same time: "-remap Super_R-Mode_switch,Menu-Multi_key" or use "-remap filename" to specify remappings from a file. Q-48: I have a Sun machine I run x11vnc on. Its Sun keyboard has just one Alt key labelled "Alt" and two Meta keys labelled with little diamonds. The machine where I run the VNC viewer only has Alt keys. How can I send a Meta keypress? (e.g. emacs needs this) Here are a couple ideas. The first one is to simply use xmodmap(1) to adjust the Sun X server. Perhaps xmodmap -e "keysym Alt_L = Meta_L Alt_L" will do the trick. (there are other ways to do it, one user used: xmodmap -e "keycode 26 = Meta_L" for his setup). Since xmodmap(1) modifies the X server mappings you may not want to do this (because it affects local work on that machine). Something like the -remap Alt_L-Meta_L to x11vnc may be sufficient for ones needs, and does not modify the X server environment. Note that you cannot send Alt_L in this case, maybe -remap Super_L-Meta_L would be a better choice. Q-49: Can I map a keystroke to a mouse button click on the remote machine? This can be done directly in some X servers using AccessX and Pointer_EnableKeys, but is a bit awkward. It may be more convenient to have x11vnc do the remapping. This can be done via the -remap option using the fake "keysyms" Button1, Button2, etc. as the "to" keys (i.e. the ones after the "-") As an example, consider a laptop that has a touchpad with only two buttons. It is difficult to do a middle button "paste" because (using XFree86 Emulate3Buttons) you have to click both buttons on the touch pad at the same time. This remapping: -remap Super_R-Button2 maps the Super_R "flag" key press to the Button2 click, thereby making pasting a bit easier. Note that once the key goes down, the button down and button up events are generated immediately on the x11vnc side. When the key is released (i.e. goes up) no events are generated. Q-50: The remote display is larger (in number of pixels) than the local display I am running the vncviewer on. I don't like the vncviewer scrollbars, what I can do? vncviewer has a option (usually accessible via F8 key or -fullscreen option) for vncviewer to run in full screen, where it will automatically scroll when the mouse is near the edge of the current view. For quick scrolling, also make sure Backing Store is enabled on the machine vncviewer is run on. (XFree86 disables it by default for some reason, add Option "backingstore" to XF86Config on the vncviewer side). BTW, contact me if you are having problems with vncviewer in fullscreen mode with your window manager (i.e. no keyboard response). I have a workaround for vncviewer using XGrabServer(). There may also be scaling viewers out there (e.g. TightVNC on Windows) that automatically shrink or expand the remote framebuffer to fit the local display. Especially for hand-held devices. See also [141]this FAQ Q-51: Does x11vnc support server-side framebuffer scaling? (E.g. to make the desktop smaller). As of Jun/2004 in the libvncserver CVS x11vnc provides basic server-side scaling. It is a global scaling of the desktop, not a per-client setting. To enable it use the "-scale fraction" option. "fraction" can either be a floating point number (e.g. -scale 0.5) or the alternative m/n fraction notation (e.g. -scale 2/3). The current implementation is experimental in that it still needs to be optimized. Extra resources (CPU, memory I/O, and memory) are required to do the scaling. If the machine is slow where x11vnc is run with scaling enabled, the interactive response can be unacceptable. OTOH, if run with scaling on a fast machine the performance degradation is usually not a big issue or even noticeable. Also, if you just want a quick, rough "thumbnail" of the display you can append ":nb" to the fraction to turn on "no blending" mode. E.g.: "-scale 1/3:nb" Fonts will be difficult to read, but the larger features will be recognizable. BTW, "no blending" mode is forced on when scaling 8bpp PseudoColor displays (because blending an indexed colormap is a bad idea and leads to random colors). In general for a scaled display if you are using a TightVNC viewer you may want to turn off jpeg encoding (e.g. vncviewer -nojpeg host:0). There appears to be a noise enhancement effect, especially for regions containing font/text: the scaling can introduce some pixel artifacts that evidently causes the TightVNC to incorrectly detect the regions as image data and thereby introduce additional pixel artifacts due to the lossiness of the jpeg compression algorithm. Experiment to see if -nojpeg improves the readability of text when using -scale to shrink the display size. Also note that scaling may actually slow down the transfer of text regions because after being scaled they do not compress as well. Another issue is that it appears VNC viewers require the screen width to be a multiple of 4. When scaling x11vnc will round the width to the nearest multiple of 4. To disable this use the ":n4" sub option (like ":nb" in the previous paragraph; to specify both use a comma: ":nb,n4", etc.) If one desires per-client scaling for something like 1:1 from a workstation and 1:2 from a smaller device (e.g. handheld), currently the only option is to run two (or more) x11vnc processes with different scalings listening on separate ports (-rfbport option, etc.). Q-52: Does x11vnc work with Xinerama? (i.e. multiple monitors joined together to form one big, single screen). Yes, it should generally work because it simply polls the big effective screen. If the viewing-end monitor is not as big as the remote Xinerama display, then the vncviewer scrollbars, etc, will have to be used to pan across the large area. There are a couple potential issues however. If the screen is not rectangular (e.g. 1280x1024 and 1024x768 monitors joined together), then there will be "non-existent" areas on the screen. The X server will return "garbage" image data for these areas and so they may be distracting to the viewer. The -blackout x11vnc option allows you to blacken-out rectangles by specifying their WxH+X+Y geometries. If your system has the libXinerama library, the -xinerama x11vnc option can be used to have it automatically determine the rectangles to be blackened out. (Note on 8bpp PseudoColor displays the fill color may not be black). Some users have reported that the mouse does not behave properly for their Xinerama display: i.e. the mouse cannot be moved to all regions of the large display. If this happens try using the -xwarppointer option. This instructs x11vnc to fake mouse pointer motions using the XWarpPointer function instead of the XTestFakeMotionEvent XTEST function. (This may be due to a bug in the X server for XTEST when Xinerama is enabled). Q-53: Can I use x11vnc on a multi-headed display that is not Xinerama (i.e. separate screens :0.0, :0.1, ... for each monitor)? You can, but it is a little bit awkward: you must start separate x11vnc processes for each screen, and on the viewing end start up separate VNC viewer processes connecting to them. e.g. on the remote end: x11vnc -display :0.0 -bg -q -rfbport 5900 x11vnc -display :0.1 -bg -q -rfbport 5901 (this could be automated in the display manager Xsetup for example) and then on the local machine where you are sitting: vncviewer somehost:0 & vncviewer somehost:1 & Note: if you are running on Solaris 8 or earlier you can easily hit up against the maximum of 6 shm segments per process (for Xsun in this case) from running multiple x11vnc processes. You should modify /etc/system as mentioned in another [142]FAQ to increase the limit. It is probably also a good idea to run with the -onetile option in this case (to limit each x11vnc to 3 shm segments), or even -noshm to use no shm segments. Q-54: Does x11vnc support the XRANDR (X Resize, Rotate and Reflection) extension? Whenever I rotate or resize the screen x11vnc just seems to crash. As of Dec/2004 in the libvncserver CVS x11vnc supports XRANDR. You enable it with the -xrandr option to make x11vnc monitor XRANDR events and also trap X server errors if the screen change occurred in the middle of an X call like XGetImage. Once it traps the screen change it will create a new framebuffer using the new screen. If the connected vnc viewers support the NewFBSize VNC extension (Windows TightVNC viewer and RealVNC 4.0 windows and Unix viewers do) then the viewer will automatically resize. Otherwise, the new framebuffer is fit as best as possible into the original viewer size (portions of the screen may be clipped, unused, etc). If you specify "-xrandr newfbsize" then vnc viewers that do not support NewFBSize will be disconnected before the resize. If you specify "-xrandr exit" then all will be disconnected and x11vnc will terminate. Q-55: Why is the view in my VNC viewer completely black? Or why is everything flashing around randomly? See the next FAQ for a possible explanation. Q-56: I use Linux Virtual Consoles (VC's) to implement 'Fast User Switching' between users' sessions (e.g. Betty is on Ctrl-Alt-F7, Bobby is on Ctrl-Alt-F8, and Sid is on Ctrl-Alt-F1: they use those keystrokes to switch between their sessions). How come the view in a VNC viewer connecting to x11vnc is either completely black or otherwise all messed up unless the X session x11vnc is attached to is in the active VC? This seems to have to do with how applications (the X server processes in this case) must "play nicely" if they are not on the active VC. That is, they should not read from the keyboard or mouse or manage the video display unless they have the active VC. Given that it appears the XGetImage() call must ultimately retrieve the framebuffer data from the video hardware itself, it would make sense x11vnc's polling wouldn't work unless the X session had active control of the VC. There does not seem to be an easy way to work around this. Even xwd(1) doesn't work in this case (try it). Something would need to be done at a lower level, say in the XFree86 X server. Also, using the XFree86 Shadow Framebuffer (a copy of the video framebuffer is kept in main memory) does not appear to fix the problem. If no one is sitting at the workstation and you just want to remotely switch the VC over to the one associated with your X session (so x11vnc can poll it), one can use the switchto(1) command, e.g. "switchto 7" for VC #7. Q-57: Does the Clipboard/Selection get transferred between the vncviewer and the X display? As of Jan/2004 in the libvncserver CVS x11vnc supports the "CutText" part of the rfb protocol. Furthermore, x11vnc is able to hold the PRIMARY selection (Xvnc does not seem to do this). If you don't want the Clipboard/Selection exchanged use the -nosel option. If you don't want the PRIMARY selection to be polled for changes use the -noprimary option. Q-58: Why don't I hear the "Beeps" in my X session (e.g. when typing tput bel in an xterm)? As of Dec/2003 in the libvncserver CVS "Beep" XBell events are tracked by default. The X server must support the XKEYBOARD extension (this is not on by default in Solaris, see Xserver(1) for how to turn it on), and so you won't hear them if the extension is not present. If you don't want to hear the beeps use the -nobell option. If you want to hear the audio from the remote applications, consider trying a redirector such as esd. Q-59: I am using x11vnc where my local machine has "popup/hidden taskbars" (e.g. GNOME or MacOS X) and the remote display where x11vnc runs also has "popup/hidden taskbars" (e.g. GNOME). When I move the mouse to the edge of the screen where the popups happen, the taskbars interfere and fight with each other in strange ways. What can I do? Is there a way to temporarily disable one or both of these magic desktop taskbars? One x11vnc user suggests: it should be straightforward to right mouse click on the task bar panel, and uncheck "enable auto-hide" from the panel properties dialog box. This will make the panel always visible. References 1. http://www.karlrunge.com/x11vnc/index.html#faq 2. http://www.karlrunge.com/x11vnc/index.html#downloading 3. http://www.karlrunge.com/x11vnc/index.html#building 4. http://www.karlrunge.com/x11vnc/index.html#faq 5. http://www.karlrunge.com/x11vnc/index.html#contact 6. http://www.uk.research.att.com/vnc/ 7. http://www.realvnc.com/ 8. http://www.tightvnc.com/ 9. http://www.karlrunge.com/x11vnc/index.html#downloading 10. http://www.tightvnc.com/download.html 11. http://www.karlrunge.com/x11vnc/index.html#xperms 12. http://www.karlrunge.com/x11vnc/index.html#viewer-download 13. http://www.sun.com/software/solaris/freeware/ 14. http://www.karlrunge.com/x11vnc/x11vnc_opts.html 15. http://www.karlrunge.com/x11vnc/index.html#accept 16. http://www.karlrunge.com/x11vnc/index.html#passwd 17. http://www.karlrunge.com/x11vnc/index.html#passwdfile 18. http://www.karlrunge.com/x11vnc/index.html#allow_opt 19. http://www.karlrunge.com/x11vnc/index.html#tcp_wrappers 20. http://sourceforge.net/projects/libvncserver/ 21. http://sourceforge.net/project/showfiles.php?group_id=32584&package_id=119006&release_id=257442 22. http://www.karlrunge.com/x11vnc/x11vnc.c 23. http://www.karlrunge.com/x11vnc/index.html#binaries 24. http://www.tightvnc.com/download.html 25. http://www.realvnc.com/download-free.html 26. http://sourceforge.net/projects/cotvnc/ 27. http://www.karlrunge.com/x11vnc/rx11vnc 28. http://www.karlrunge.com/x11vnc/rx11vnc.pl 29. ftp://ftp.uu.net/graphics/jpeg/ 30. http://www.gzip.org/zlib/ 31. http://www.sunfreeware.com/ 32. http://www.karlrunge.com/x11vnc/index.html#solaris251build 33. http://www.tightvnc.com/ 34. http://www.karlrunge.com/x11vnc/x11vnc_opts.html 35. http://www.karlrunge.com/x11vnc/index.html#passwd 36. http://www.karlrunge.com/x11vnc/recurse_x11vnc.jpg 37. http://wwws.sun.com/sunray/index.html 38. mailto:xvml@karlrunge.com 39. http://www.karlrunge.com/x11vnc/index.html#FAQ-1 40. http://www.karlrunge.com/x11vnc/index.html#FAQ-2 41. http://www.karlrunge.com/x11vnc/index.html#FAQ-3 42. http://www.karlrunge.com/x11vnc/index.html#FAQ-4 43. http://www.karlrunge.com/x11vnc/index.html#FAQ-5 44. http://www.karlrunge.com/x11vnc/index.html#FAQ-6 45. http://www.karlrunge.com/x11vnc/index.html#FAQ-7 46. http://www.karlrunge.com/x11vnc/index.html#FAQ-8 47. http://www.karlrunge.com/x11vnc/index.html#FAQ-9 48. http://www.karlrunge.com/x11vnc/index.html#FAQ-10 49. http://www.karlrunge.com/x11vnc/index.html#FAQ-11 50. http://www.karlrunge.com/x11vnc/index.html#FAQ-12 51. http://www.karlrunge.com/x11vnc/index.html#FAQ-13 52. http://www.karlrunge.com/x11vnc/index.html#FAQ-14 53. http://www.karlrunge.com/x11vnc/index.html#FAQ-15 54. http://www.karlrunge.com/x11vnc/index.html#FAQ-16 55. http://www.karlrunge.com/x11vnc/index.html#FAQ-17 56. http://www.karlrunge.com/x11vnc/index.html#FAQ-18 57. http://www.karlrunge.com/x11vnc/index.html#FAQ-19 58. http://www.karlrunge.com/x11vnc/index.html#FAQ-20 59. http://www.karlrunge.com/x11vnc/index.html#FAQ-21 60. http://www.karlrunge.com/x11vnc/index.html#FAQ-22 61. http://www.karlrunge.com/x11vnc/index.html#FAQ-23 62. http://www.karlrunge.com/x11vnc/index.html#FAQ-24 63. http://www.karlrunge.com/x11vnc/index.html#FAQ-25 64. http://www.karlrunge.com/x11vnc/index.html#FAQ-26 65. http://www.karlrunge.com/x11vnc/index.html#FAQ-27 66. http://www.karlrunge.com/x11vnc/index.html#FAQ-28 67. http://www.karlrunge.com/x11vnc/index.html#FAQ-29 68. http://www.karlrunge.com/x11vnc/index.html#FAQ-30 69. http://www.karlrunge.com/x11vnc/index.html#FAQ-31 70. http://www.karlrunge.com/x11vnc/index.html#FAQ-32 71. http://www.karlrunge.com/x11vnc/index.html#FAQ-33 72. http://www.karlrunge.com/x11vnc/index.html#FAQ-34 73. http://www.karlrunge.com/x11vnc/index.html#FAQ-35 74. http://www.karlrunge.com/x11vnc/index.html#FAQ-36 75. http://www.karlrunge.com/x11vnc/index.html#FAQ-37 76. http://www.karlrunge.com/x11vnc/index.html#FAQ-38 77. http://www.karlrunge.com/x11vnc/index.html#FAQ-39 78. http://www.karlrunge.com/x11vnc/index.html#FAQ-40 79. http://www.karlrunge.com/x11vnc/index.html#FAQ-41 80. http://www.karlrunge.com/x11vnc/index.html#FAQ-42 81. http://www.karlrunge.com/x11vnc/index.html#FAQ-43 82. http://www.karlrunge.com/x11vnc/index.html#FAQ-44 83. http://www.karlrunge.com/x11vnc/index.html#FAQ-45 84. http://www.karlrunge.com/x11vnc/index.html#FAQ-46 85. http://www.karlrunge.com/x11vnc/index.html#FAQ-47 86. http://www.karlrunge.com/x11vnc/index.html#FAQ-48 87. http://www.karlrunge.com/x11vnc/index.html#FAQ-49 88. http://www.karlrunge.com/x11vnc/index.html#FAQ-50 89. http://www.karlrunge.com/x11vnc/index.html#FAQ-51 90. http://www.karlrunge.com/x11vnc/index.html#FAQ-52 91. http://www.karlrunge.com/x11vnc/index.html#FAQ-53 92. http://www.karlrunge.com/x11vnc/index.html#FAQ-54 93. http://www.karlrunge.com/x11vnc/index.html#FAQ-55 94. http://www.karlrunge.com/x11vnc/index.html#FAQ-56 95. http://www.karlrunge.com/x11vnc/index.html#FAQ-57 96. http://www.karlrunge.com/x11vnc/index.html#FAQ-58 97. http://www.karlrunge.com/x11vnc/index.html#FAQ-59 98. http://www.karlrunge.com/x11vnc/index.html#solarisbuilding 99. http://www.karlrunge.com/x11vnc/index.html#building 100. http://www.karlrunge.com/x11vnc/index.html#buildfaq 101. http://packages.debian.org/x11vnc 102. http://www.linuxpackages.net/search_view.php?by=name&name=x11vnc 103. http://dag.wieers.com/packages/x11vnc/ 104. http://www.bell-labs.com/project/wwexptools/packages.html 105. http://www.karlrunge.com/x11vnc/index.html#solarisbuilding 106. http://www.tightvnc.com/download.html 107. http://www.realvnc.com/download-free.html 108. http://sourceforge.net/projects/cotvnc/ 109. http://www.karlrunge.com/x11vnc/x11vnc_opts.html 110. http://www.karlrunge.com/x11vnc/index.html#tunnelling 111. http://www.karlrunge.com/x11vnc/index.html#passwd 112. http://www.karlrunge.com/x11vnc/index.html#passwdfile 113. http://fredrik.hubbe.net/x2vnc.html 114. http://www.hubbe.net/~hubbe/win2vnc.html 115. http://www.deboer.gmxhome.de/ 116. http://sourceforge.net/projects/win2vnc/ 117. http://fredrik.hubbe.net/x2vnc.html 118. http://freshmeat.net/projects/x2x/ 119. http://ftp.digital.com/pub/Digital/SRC/x2x/ 120. http://zapek.com/software/zvnc/ 121. http://www.karlrunge.com/x11vnc/index.html#8bpp 122. http://www.karlrunge.com/x11vnc/index.html#overlays 123. http://www.karlrunge.com/x11vnc/index.html#inetd 124. ftp://ftp.x.org/ 125. http://www.karlrunge.com/x11vnc/dtVncPopup 126. http://www.karlrunge.com/x11vnc/index.html#passwdfile 127. http://www.karlrunge.com/x11vnc/index.html#tunnelling 128. http://www.karlrunge.com/x11vnc/index.html#tunnelling 129. http://www.karlrunge.com/x11vnc/index.html#tunnelling 130. http://www.karlrunge.com/x11vnc/index.html#tunnelling 131. http://www.karlrunge.com/x11vnc/blockdpy.c 132. http://www.karlrunge.com/x11vnc/x11vnc.c 133. http://www.karlrunge.com/x11vnc/x11vnc_loop 134. http://www.tightvnc.com/ 135. http://www.karlrunge.com/x11vnc/shm_clear 136. http://www.karlrunge.com/x11vnc/index.html#noshm 137. http://www.tightvnc.com/ 138. http://www.karlrunge.com/x11vnc/index.html#greaterless 139. http://www.karlrunge.com/x11vnc/index.html#xkbmodtweak 140. http://www.karlrunge.com/x11vnc/index.html#xkbmodtweak 141. http://www.karlrunge.com/x11vnc/index.html#scaling 142. http://www.karlrunge.com/x11vnc/index.html#solshm ======================================================================= http://www.karlrunge.com/x11vnc/x11vnc_opts.html: _________________________________________________________________ x11vnc: a VNC server for real X displays Here are all of x11vnc command line options: % x11vnc -help x11vnc: allow VNC connections to real X11 displays. 0.7pre lastmod: 2004-12-20 Typical usage is: Run this command in a shell on the remote machine "far-host" with X session you wish to view: x11vnc -display :0 Then run this in another window on the machine you are sitting at: vncviewer far-host:0 Once x11vnc establishes connections with the X11 server and starts listening as a VNC server it will print out a string: PORT=XXXX where XXXX is typically 5900 (the default VNC server port). One would next run something like this on the local machine: "vncviewer hostname:N" where "hostname" is the name of the machine running x11vnc and N is XXXX - 5900, i.e. usually "vncviewer hostname:0". By default x11vnc will not allow the screen to be shared and it will exit as soon as a client disconnects. See -shared and -forever below to override these protections. See the FAQ on how to tunnel the VNC connection through an encrypted channel such as ssh(1). For additional info see: http://www.karlrunge.com/x11vnc/ and http://www.karlrunge.com/x11vnc/#faq Rudimentary config file support: if the file $HOME/.x11vncrc exists then each line in it is treated as a single command line option. Disable with -norc. For each option name, the leading character "-" is not required. E.g. a line that is either "nap" or "-nap" may be used and are equivalent. Likewise "wait 100" or "-wait 100" are acceptable and equivalent lines. The "#" character comments out to the end of the line in the usual way. Leading and trailing whitespace is trimmed off. Lines may be continued with a "\" as the last character of a line (it becomes a space character). Options: -display disp X11 server display to connect to, usually :0. The X server process must be running on same machine and support MIT-SHM. Equivalent to setting the DISPLAY environment variable to "disp". -auth file Set the X authority file to be "file", equivalent to setting the XAUTHORITY environment varirable to "file" before startup. See Xsecurity(7), xauth(1) man pages. -id windowid Show the window corresponding to "windowid" not the entire display. New windows like popup menus, transient toplevels, etc, may not be seen or may be clipped. Disabling SaveUnders or BackingStore in the X server may help show them. x11vnc may crash if the window is initially partially obscured, changes size, is iconified, etc. Some steps are taken to avoid this and the -xrandr mechanism is used to track resizes. Use xwininfo(1) to get the window id, or use "-id pick" to have x11vnc run xwininfo(1) for you and extract the id. The -id option is useful for exporting very simple applications (e.g. the current view on a webcam). -sid windowid As -id, but instead of using the window directly it shifts a root view to it: this shows SaveUnders menus, etc, although they will be clipped if they extend beyond the window. -flashcmap In 8bpp indexed color, let the installed colormap flash as the pointer moves from window to window (slow). -notruecolor For 8bpp displays, force indexed color (i.e. a colormap) even if it looks like 8bpp TrueColor. (rare problem) -visual n Experimental option: probably does not do what you think. It simply *forces* the visual used for the framebuffer; this may be a bad thing... (e.g. messes up colors or cause a crash). It is useful for testing and for some workarounds. n may be a decimal number, or 0x hex. Run xdpyinfo(1) for the values. One may also use "TrueColor", etc. see for a list. If the string ends in ":m" for better or for worse the visual depth is forced to be m. -overlay Handle multiple depth visuals on one screen, e.g. 8+24 and 24+8 overlay visuals (the 32 bits per pixel are packed with 8 for PseudoColor and 24 for TrueColor). Currently -overlay only works on Solaris via XReadScreen(3X11) and IRIX using XReadDisplay(3). On Solaris there is a problem with image "bleeding" around transient popup menus (but not for the menu itself): a workaround is to disable SaveUnders by passing the "-su" argument to Xsun (in /etc/dt/config/Xservers). Also note that the mouse cursor shape is exactly correct in this mode. Use -overlay as a workaround for situations like these: Some legacy applications require the default visual to be 8bpp (8+24), or they will use 8bpp PseudoColor even when the default visual is depth 24 TrueColor (24+8). In these cases colors in some windows will be messed up in x11vnc unless -overlay is used. Another use of -overlay is to enable showing the exact mouse cursor shape (details below). Under -overlay, performance will be somewhat degraded due to the extra image transformations required. For optimal performance do not use -overlay, but rather configure the X server so that the default visual is depth 24 TrueColor and try to have all apps use that visual (some apps have -use24 or -visual options). -overlay_nocursor Sets -overlay, but does not try to draw the exact mouse cursor shape using the overlay mechanism. -scale fraction Scale the framebuffer by factor "fraction". Values less than 1 shrink the fb, larger ones expand it. Note: image may not be sharp and response may be slower. Currently the cursor shape is not scaled. If "fraction" contains a decimal point "." it is taken as a floating point number, alternatively the notation "m/n" may be used to denote fractions exactly, e.g. -scale 2/3. Scaling Options: can be added after "fraction" via ":", to supply multiple ":" options use commas. If you just want a quick, rough scaling without blending, append ":nb" to "fraction" (e.g. -scale 1/3:nb). For compatibility with vncviewers the scaled width is adjusted to be a multiple of 4: to disable this use ":n4". More esoteric options: ":in" use interpolation scheme even when shrinking, ":pad", pad scaled width and height to be multiples of scaling denominator (e.g. 3 for 2/3). -viewonly All VNC clients can only watch (default off). -shared VNC display is shared (default off). -once Exit after the first successfully connected viewer disconnects, opposite of -forever. This is the Default. -forever Keep listening for more connections rather than exiting as soon as the first client(s) disconnect. Same as -many -inetd Launched by inetd(1): stdio instead of listening socket. Note: if you are not redirecting stderr to a log file (via shell 2> or -o option) you must also specify the -q option, otherwise the stderr goes to the viewer. -connect string For use with "vncviewer -listen" reverse connections. If "string" has the form "host" or "host:port" the connection is made once at startup. Use commas for a list of host's and host:port's. If "string" contains "/" it is instead interpreted as a file to periodically check for new hosts. The first line is read and then the file is truncated. -vncconnect Monitor the VNC_CONNECT X property set by the standard -novncconnect VNC program vncconnect(1). When the property is set to "host" or "host:port" establish a reverse connection. Using xprop(1) instead of vncconnect may work (see the FAQ). Default: -vncconnect -allow host1[,host2..] Only allow client connections from hosts matching the comma separated list of hostnames or IP addresses. Can also be a numerical IP prefix, e.g. "192.168.100." to match a simple subnet, for more control build libvncserver with libwrap support (See the FAQ). If the list contains a "/" it instead is a interpreted as a file containing addresses or prefixes that is re-read each time a new client connects. Lines can be commented out with the "#" character in the usual way. -localhost Same as -allow 127.0.0.1 -viewpasswd string Supply a 2nd password for view-only logins. The -passwd (full-access) password must also be supplied. -passwdfile filename Specify libvncserver -passwd via the first line of the file "filename" instead of via command line. If a second non blank line exists in the file it is taken as a view-only password (i.e. -viewpasswd) Note: this is a simple plaintext passwd, see also -rfbauth and -storepasswd below for obfuscated passwords. -storepasswd pass file Store password "pass" as the VNC password in the file "file". Once the password is stored the program exits. Use the password via "-rfbauth file" -accept string Run a command (possibly to prompt the user at the X11 display) to decide whether an incoming client should be allowed to connect or not. "string" is an external command run via system(3) or some special cases described below. Be sure to quote "string" if it contains spaces, shell characters, etc. If the external command returns 0 the client is accepted, otherwise the client is rejected. See below for an extension to accept a client view-only. Environment: The RFB_CLIENT_IP environment variable will be set to the incoming client IP number and the port in RFB_CLIENT_PORT (or -1 if unavailable). Similarly, RFB_SERVER_IP and RFB_SERVER_PORT (the x11vnc side of the connection), are set to allow identification of the tcp virtual circuit. The x11vnc process id will be in RFB_X11VNC_PID, a client id number in RFB_CLIENT_ID, and the number of other connected clients in RFB_CLIENT_COUNT. RFB_MODE will be "accept" If "string" is "popup" then a builtin popup window is used. The popup will time out after 120 seconds, use "popup:N" to modify the timeout to N seconds (use 0 for no timeout) If "string" is "xmessage" then an xmessage(1) invocation is used for the command. xmessage must be installed on the machine for this to work. Both "popup" and "xmessage" will present an option for accepting the client "View-Only" (the client can only watch). This option will not be presented if -viewonly has been specified, in which case the entire display is view only. If the user supplied command is prefixed with something like "yes:0,no:*,view:3 mycommand ..." then this associates the numerical command return code with the actions: accept, reject, and accept-view-only, respectively. Use "*" instead of a number to indicate the default action (in case the command returns an unexpected value). E.g. "no:*" is a good choice. Note that x11vnc blocks while the external command or popup is running (other clients may see no updates during this period). More -accept tricks: use "popupmouse" to only allow mouse clicks in the builtin popup to be recognized. Similarly use "popupkey" to only recognize keystroke responses. These are to help avoid the user accidentally accepting a client by typing or clicking. All 3 of the popup keywords can be followed by +N+M to supply a position for the popup window. The default is to center the popup window. -gone string As -accept, except to run a user supplied command when a client goes away (disconnects). RFB_MODE will be set to "gone" and the other RFB_* variables are as in -accept. Unlike -accept, the command return code is not interpreted by x11vnc. Example: -gone 'xlock &' -noshm Do not use the MIT-SHM extension for the polling. Remote displays can be polled this way: be careful this can use large amounts of network bandwidth. This is also of use if the local machine has a limited number of shm segments and -onetile is not sufficient. -flipbyteorder Sometimes needed if remotely polled host has different endianness. Ignored unless -noshm is set. -onetile Do not use the new copy_tiles() framebuffer mechanism, just use 1 shm tile for polling. Limits shm segments used to 3. -blackout string Black out rectangles on the screen. "string" is a comma separated list of WxH+X+Y type geometries for each rectangle. -xinerama If your screen is composed of multiple monitors glued together via XINERAMA, and that screen is non-rectangular this option will try to guess the areas to black out (if your system has libXinerama). In general on XINERAMA displays you may need to use the -xwarppointer option if the mouse pointer misbehaves. -xrandr [mode] If the display supports the XRANDR (X Resize, Rotate and Reflection) extension, and you expect XRANDR events to occur to the display while x11vnc is running, this options indicates x11vnc should try to respond to them (as opposed to simply crashing by assuming the old screen size). See the xrandr(1) manpage and run 'xrandr -q' for more info. [mode] is optional and described below. Since watching for XRANDR events and errors increases polling overhead, only use this option if XRANDR changes are expected. For example on a rotatable screen PDA or laptop, or using a XRANDR-aware Desktop where you resize often. It is best to be viewing with a vncviewer that supports the NewFBSize encoding, since it knows how to react to screen size changes. Otherwise, libvncserver tries to do so something reasonable for viewers that cannot do this (portions of the screen may be clipped, unused, etc). "mode" defaults to "resize", which means create a new, resized, framebuffer and hope all viewers can cope with the change. "newfbsize" means first disconnect all viewers that do not support the NewFBSize VNC encoding, and then resize the framebuffer. "exit" means disconnect all viewer clients, and then terminate x11vnc. -padgeom WxH Whenever a new vncviewer connects, the framebuffer is replaced with a fake, solid black one of geometry WxH. Shortly afterwards the framebuffer is replaced with the real one. This is intended for use with vncviewers that do not support NewFBSize and one wants to make sure the initial viewer geometry will be big enough to handle all subsequent resizes (e.g. under -xrandr, -remote id:windowid, rescaling, etc. -o logfile Write stderr messages to file "logfile" instead of to the terminal. Same as "-logfile file". -rc filename Use "filename" instead of $HOME/.x11vncrc for rc file. -norc Do not process any .x11vncrc file for options. -h, -help Print this help text. -V, -version Print program version (last modification date). -q Be quiet by printing less informational output to stderr. Same as -quiet. -bg Go into the background after screen setup. Messages to stderr are lost unless -o logfile is used. Something like this could be useful in a script: port=`ssh $host "x11vnc -display :0 -bg" | grep PORT` port=`echo "$port" | sed -e 's/PORT=//'` port=`expr $port - 5900` vncviewer $host:$port -modtweak Option -modtweak automatically tries to adjust the AltGr -nomodtweak and Shift modifiers for differing language keyboards between client and host. Otherwise, only a single key press/release of a Keycode is simulated (i.e. ignoring the state of the modifiers: this usually works for identical keyboards). Also useful in resolving cases where a Keysym is bound to multiple keys (e.g. "<" + ">" and "," + "<" keys). Default: -modtweak -xkb When in modtweak mode, use the XKEYBOARD extension (if the X display supports it) to do the modifier tweaking. This is powerful and should be tried if there are still keymapping problems when using -modtweak by itself. -skip_keycodes string Ignore the comma separated list of decimal keycodes. Perhaps these are keycodes not on your keyboard but your X server thinks exist. Currently only applies to -xkb mode. Use this option to help x11vnc in the reverse problem it tries to solve: Keysym -> Keycode(s) when ambiguities exist (more than one Keycode per Keysym). Run 'xmodmap -pk' to see your keymapping. E.g. "-skip_keycodes 94,114" -add_keysyms If a Keysym is received from a VNC viewer and that Keysym does not exist in the X server, then add the Keysym to the X server's keyboard mapping. Added Keysyms will be removed when exiting. -clear_mods At startup and exit clear the modifier keys by sending KeyRelease for each one. The Lock modifiers are skipped. Used to clear the state if the display was accidentally left with any pressed down. -clear_keys As -clear_mods, except try to release any pressed key. Note that this option and -clear_mods can interfere with a person typing at the physical keyboard. -remap string Read Keysym remappings from file named "string". Format is one pair of Keysyms per line (can be name or hex value) separated by a space. If no file named "string" exists, it is instead interpreted as this form: key1-key2,key3-key4,... See header file for a list of Keysym names, or use xev(1). To map a key to a button click, use the fake Keysyms "Button1", ..., etc. E.g. "-remap Super_R-Button2" (useful for pasting on a laptop) -norepeat Option -norepeat disables X server key auto repeat -repeat when VNC clients are connected. This works around a repeating keystrokes bug (triggered by long processing delays between key down and key up client events: either from large screen changes or high latency). Note: your VNC viewer side will likely do autorepeating, so this is no loss unless someone is simultaneously at the real X display. Default: -norepeat -nofb Ignore video framebuffer: only process keyboard and pointer. Intended for use with Win2VNC and x2vnc dual-monitor setups. -nobell Do not watch for XBell events. (no beeps will be heard) Note: XBell monitoring requires the XKEYBOARD extension. -nosel Do not manage exchange of X selection/cutbuffer between VNC viewers and the X server. -noprimary Do not poll the PRIMARY selection for changes to send back to clients. (PRIMARY is still set on received changes, however). -cursor [mode] Sets how the pointer cursor shape (little icon at the -nocursor mouse pointer) should be handled. The "mode" string is optional and is described below. The default is to show some sort of cursor shape(s). How this is done depends on the VNC viewer and the X server. Use -nocursor to disable cursor shapes completely. Some VNC viewers support the TightVNC CursorPosUpdates and CursorShapeUpdates extensions (cuts down on network traffic by not having to send the cursor image every time the pointer is moved), in which case these extensions are used (see -nocursorshape and -nocursorpos below to disable). For other viewers the cursor shape is written directly to the framebuffer every time the pointer is moved or changed and gets sent along with the other framebuffer updates. In this case, there will be some lag between the vnc viewer pointer and the remote cursor position. If the X display supports retrieving the cursor shape information from the X server, then the default is to use that mode. On Solaris this can be done with the SUN_OVL extension using -overlay (see also the -overlay_nomouse option). A similar overlay scheme is used on IRIX. Xorg (e.g. Linux) and recent Solaris Xsun servers support the XFIXES extension to retrieve the exact cursor shape from the X server. If XFIXES is present it is preferred over Overlay and is used by default (see -noxfixes below). This can be disabled with -nocursor, and also some values of the "mode" option below. The "mode" string can be used to fine-tune the displaying of cursor shapes. It can be used the following ways: "-cursor arrow" - just show the standard arrow nothing more or nothing less. "-cursor none" - same as "-nocursor" "-cursor X" - when the cursor appears to be on the root window, draw the familiar X shape. Some desktops such as GNOME cover up the root window completely, and so this will not work, try "X1", etc, to try to shift the tree depth. On high latency links or slow machines there will be a time lag between expected and the actual cursor shape. "-cursor some" - like "X" but use additional heuristics to try to guess if the window should have a windowmanager-like resizer cursor or a text input I-beam cursor. This is a complete hack, but may be useful in some situations because it provides a little more feedback about the cursor shape. "-cursor most" - try to show as many cursors as possible. Often this will only be the same as "some" unless the display has overlay visuals or XFIXES extensions available. On Solaris and IRIX if XFIXES is not available, -overlay mode will be attempted. -noxfixes Do not use the XFIXES extension to draw the exact cursor shape even if it is available. -nocursorshape Do not use the TightVNC CursorShapeUpdates extension even if clients support it. See -cursor above. -cursorpos Option -cursorpos enables sending the X cursor position -nocursorpos back to all vnc clients that support the TightVNC CursorPosUpdates extension. Other clients will be able to see the pointer motions. Default: -cursorpos -xwarppointer Move the pointer with XWarpPointer(3X) instead of the XTEST extension. Use this as a workaround if the pointer motion behaves incorrectly, e.g. on touchscreens or other non-standard setups. Also sometimes needed on XINERAMA displays. -buttonmap string String to remap mouse buttons. Format: IJK-LMN, this maps buttons I -> L, etc., e.g. -buttonmap 13-31 Button presses can also be mapped to keystrokes: replace a button digit on the right of the dash with :: or :+: etc. for multiple keys. For example, if the viewing machine has a mouse-wheel (buttons 4 5) but the x11vnc side does not, these will do scrolls: -buttonmap 12345-123:Prior::Next: -buttonmap 12345-123:Up+Up+Up::Down+Down+Down: See header file for a list of Keysyms, or use the xev(1) program. Note: mapping of button clicks to Keysyms may not work if -modtweak or -xkb is needed for the Keysym. If you include a modifier like "Shift_L" the modifier's up/down state is toggled, e.g. to send "The" use :Shift_L+t+Shift_L+h+e: (the 1st one is shift down and the 2nd one is shift up). (note: the initial state of the modifier is ignored and not reset) To include button events use "Button1", ... etc. -nodragging Do not update the display during mouse dragging events (mouse motion with a button held down). Greatly improves response on slow setups, but you lose all visual feedback for drags, text selection, and some menu traversals. It overrides any -pointer_mode setting (think of it as pointer_mode 0) -pointer_mode n Various pointer update schemes. The problem is pointer motion can cause rapid changes on the screen, e.g. a window drag. Neither x11vnc's screen polling nor the bandwidth to the vncviewers can keep up these rapid screen changes: everything bogs down when dragging or scrolling. Note that most video h/w is optimized for writing, not reading (a 50X rate difference is possible) and x11vnc is reading all the time. So a scheme has to be used to "eat" much of that pointer input before re-polling the screen. n can be 1 to 4. n=1 was the original scheme used to about Jan 2004: it basically just skips -input_skip pointer events before repolling the screen. n=2 is an improved scheme: by watching the current rate it tries to detect if it should try to "eat" more pointer events. n=3 is basically a dynamic -nodragging mode: it detects if the mouse drag motion has paused and refreshes the display. n=4 is TBD, it will try measure screen read and client write rates and try to insert "frames" between the on/off states of mode 3. The default n is 2. -input_skip n For the pointer handling when non-threaded: try to read n user input events before scanning display. n < 0 means to act as though there is always user input. Default: 10 -debug_pointer Print debugging output for every pointer event. -debug_keyboard Print debugging output for every keyboard event. Same as -dp and -dk, respectively. Use multiple times for more output. -defer time Time in ms to wait for updates before sending to client (deferUpdateTime) Default: 30 -wait time Time in ms to pause between screen polls. Used to cut down on load. Default: 30 -nap Monitor activity and if low take longer naps between polls to really cut down load when idle. Default: off -sb time Time in seconds after NO activity (e.g. screen blank) to really throttle down the screen polls (i.e. sleep for about 1.5 secs). Use 0 to disable. Default: 60 -sigpipe string Broken pipe (SIGPIPE) handling. "string" can be "ignore" or "exit". For "ignore" libvncserver will handle the abrupt loss of a client and continue, for "exit" x11vnc will cleanup and exit at the 1st broken connection. Default: "ignore". -threads Whether or not to use the threaded libvncserver -nothreads algorithm [rfbRunEventLoop] if libpthread is available Default: -nothreads -fs f If the fraction of changed tiles in a poll is greater than f, the whole screen is updated. Default: 0.75 -gaps n Heuristic to fill in gaps in rows or cols of n or less tiles. Used to improve text paging. Default: 4 -grow n Heuristic to grow islands of changed tiles n or wider by checking the tile near the boundary. Default: 3 -fuzz n Tolerance in pixels to mark a tiles edges as changed. Default: 2 -gui [gui-opts] Start up a simple tcl/tk gui based on the the remote control options -remote/-query described below. Requires the "wish" program to be installed on the machine. "gui-opts" is not required: the default is to start up both the gui and x11vnc with the gui showing up on the X display in the environment variable DISPLAY. "gui-opts" can be a comma separated list of items. Currently there are only two types of items: 1) a gui mode and 2) the X display the gui should display on. The gui mode can be "start", "conn", or "wait" "start" is the default mode above and is not required. "conn" means do not automatically start up x11vnc, but instead just try to connect to an existing x11vnc process. "wait" means just start the gui and nothing else (you will later instruct the gui to start x11vnc or connect to an existing one.) Note the possible confusion regarding the potentially two different X displays: x11vnc polls one, but you may want the gui to appear on another. For example, if you ssh in and x11vnc is not running yet you may want the gui to come back to you via your ssh redirected X display (e.g. localhost:10). Examples: "x11vnc -gui", "x11vnc -gui localhost:10", "x11vnc -gui :10", "x11vnc -gui wait,:10", "x11vnc -gui " If you do not specify a gui X display in "gui-opts" then the DISPLAY environment variable and -display option are tried (in that order). Regarding the x11vnc X display the gui will try to connect to, it first tries -display and then DISPLAY. For example, "x11vnc -display :0 -gui otherhost:0", will remote control an x11vnc polling :0 and display the gui on otherhost:0 If you do not intend to start x11vnc from the gui (i.e. just remote control an existing one), then the gui process can run on a different machine from the x11vnc server as long as X permissions, etc. permit communication between the two. -remote command Remotely control some aspects of an already running x11vnc server. "-R" and "-r" are aliases for "-remote". After the remote control command is sent to the running server the 'x11vnc -remote ...' command exits. You can often use the -query command (see below) to see if the x11vnc server processed your -remote command. The default communication channel is that of X properties (specifically VNC_CONNECT), and so this command must be run with correct settings for DISPLAY and possibly XAUTHORITY to connect to the X server and set the property. Alternatively, use the -display and -auth options to set them to the correct values. The running server cannot use the -novncconnect option because that disables the communication channel. See below for alternate channels. For example: 'x11vnc -remote stop' (which is the same as 'x11vnc -R stop') will close down the x11vnc server. 'x11vnc -R shared' will enable shared connections, and 'x11vnc -R scale:3/4' will rescale the desktop. Note: the more drastic the change induced by the -remote command, the bigger the chance for bugs or crashes. Please report reproducible bugs. The following -remote/-R commands are supported: stop terminate the server, same as "quit" "exit" or "shutdown" ping see if the x11vnc server responds. Return is: ans=ping: blacken try to push a black fb update to all clients (due to timings a client could miss it). Same as "zero", also "zero:x1,y1,x2,y2" for a rectangle. refresh send the entire fb to all clients. reset recreate the fb, polling memory, etc. id:windowid set -id window to "windowid". empty or "root" to go back to root window sid:windowid set -sid window to "windowid" flashcmap enable -flashcmap mode. noflashcmap disable -flashcmap mode. notruecolor enable -notruecolor mode. truecolor disable -notruecolor mode. overlay enable -overlay mode (if applicable). nooverlay disable -overlay mode. overlay_cursor in -overlay mode, enable cursor drawing. overlay_nocursor disable cursor drawing. same as nooverlay_cursor. visual:vis set -visual to "vis" scale:frac set -scale to "frac" viewonly enable -viewonly mode. noviewonly disable -viewonly mode. shared enable -shared mode. noshared disable -shared mode. forever enable -forever mode. noforever disable -forever mode. deny deny any new connections, same as "lock" nodeny allow new connections, same as "unlock" connect:host do reverse connection to host, "host" may be a comma separated list of hosts or host:ports. See -connect. disconnect:host disconnect any clients from "host" same as "close:host". Use host "all" to close all current clients. If you know the client internal hex ID, e.g. 0x3 (returned by -query clients and RFB_CLIENT_ID), you can use that too. allowonce:host For the next connection only, allow connection from "host". allow:hostlist set -allow list to (comma separated) "hostlist". See -allow and -localhost. Do not use with -allow /path/to/file Use "+host" to add a single host, and use "-host" to delete a single host localhost enable -localhost mode nolocalhost disable -localhost mode accept:cmd set -accept "cmd" (empty to disable). gone:cmd set -gone "cmd" (empty to disable). noshm enable -noshm mode. shm disable -noshm mode (i.e. use shm). flipbyteorder enable -flipbyteorder mode, you may need to set noshm for this to do something. noflipbyteorder disable -flipbyteorder mode. onetile enable -onetile mode. (you may need to set shm for this to do something) noonetile disable -onetile mode. blackout:str set -blackout "str" (empty to disable). See -blackout for the form of "str" (basically: WxH+X+Y,...) Use "+WxH+X+Y" to append a single rectangle use "-WxH+X+Y" to delete one xinerama enable -xinerama mode. (if applicable) noxinerama disable -xinerama mode. xrandr enable -xrandr mode. (if applicable) noxrandr disable -xrandr mode. xrandr_mode:mode set the -xrandr mode to "mode". padgeom:WxH set -padgeom to WxH (empty to disable) If WxH is "force" or "do" the padded geometry fb is immediately applied. quiet enable -quiet mode. noquiet disable -quiet mode. modtweak enable -modtweak mode. nomodtweak enable -nomodtweak mode. xkb enable -xkb modtweak mode. noxkb disable -xkb modtweak mode. skip_keycodes:str enable -xkb -skip_keycodes "str". add_keysyms enable -add_keysyms mode. noadd_keysyms stop adding keysyms. those added will still be removed at exit. clear_mods enable -clear_mods mode and clear them. noclear_mods disable -clear_mods mode. clear_keys enable -clear_keys mode and clear them. noclear_keys disable -clear_keys mode. remap:str set -remap "str" (empty to disable). See -remap for the form of "str" (basically: key1-key2,key3-key4,...) Use "+key1-key2" to append a single keymapping, use "-key1-key2" to delete. norepeat enable -norepeat mode. repeat disable -norepeat mode. bell enable bell (if supported). nobell disable bell. sel disable -nosel mode. nosel enable -nosel mode. primary disable -noprimary mode. noprimary enable -noprimary mode. cursor:mode enable -cursor "mode". show_cursor enable showing a cursor. noshow_cursor disable showing a cursor. (same as "nocursor") xfixes enable xfixes cursor shape mode. noxfixes disable xfixes cursor shape mode. cursorshape disable -nocursorshape mode. nocursorshape enable -nocursorshape mode. cursorpos disable -nocursorpos mode. nocursorpos enable -nocursorpos mode. xwarp enable -xwarppointer mode. noxwarp disable -xwarppointer mode. buttonmap:str set -buttonmap "str", empty to disable dragging disable -nodragging mode. nodragging enable -nodragging mode. pointer_mode n set -pointer_mode to n. input_skip n set -input_skip to n. debug_pointer enable -debug_pointer, same as "dp" nodebug_pointer disable -debug_pointer, same as "nodp" debug_keyboard enable -debug_keyboard, same as "dk" nodebug_keyboard disable -debug_keyboard, same as "nodk" defer:n set -defer to n ms,same as deferupdate:n wait:n set -wait to n ms. nap enable -nap mode. nonap disable -nap mode. sb:n set -sb to n s, same as screen_blank:n fs:frac set -fs fraction to "frac", e.g. 0.5 gaps:n set -gaps to n. grow:n set -grow to n. fuzz:n set -fuzz to n. progressive:n set libvncserver -progressive slice height parameter to n. file:name run -remote commands from file "name", one command per line,blank and # skipped noremote disable the -remote command processing, it cannot be turned back on. The vncconnect(1) command from standard VNC distributions may also be used if string is prefixed with "cmd=" E.g. 'vncconnect cmd=stop'. Under some circumstances xprop(1) can used if it supports -set (see the FAQ). If "-connect /path/to/file" has been supplied to the running x11vnc server then that file can be used as a communication channel (this is the only way to remote control one of many x11vnc's polling the same X display) Simply run: 'x11vnc -connect /path/to/file -remote ...' or you can directly write to the file via something like: "echo cmd=stop > /path/to/file", etc. -query variable Like -remote, except just query the value of "variable". "-Q" is an alias for "-query". Multiple queries can be done by separating variables by commas, e.g. -query var1,var2. The results come back in the form ans=var1:value1,ans=var2:value2,... to the standard output. If a variable is read-only, it comes back with prefix "aro=" instead of "ans=". Some -remote commands are pure actions that do not make sense as variables, e.g. "stop" or "disconnect", in these cases the value returned is "N/A". To direct a query straight to the VNC_CONNECT property or connect file use "qry=..." instead of "cmd=..." Here is the current list of "variables" that can be supplied to the -query command. This includes the "N/A" ones that return no useful info. For variables names that do not correspond to an x11vnc option or remote command, we hope the name makes it obvious what the returned value corresponds to (hint: the ext_* variables correspond to the presence of X extensions): ans= stop quit exit shutdown ping blacken zero refresh reset close disconnect id sid flashcmap noflashcmap truecolor notruecolor overlay nooverlay overlay_cursor overlay_yescursor nooverlay_cursor overlay_nocursor visual scale viewonly noviewonly shared noshared forever noforever once deny lock nodeny unlock connect allowonce allow localhost nolocalhost accept gone shm noshm flipbyteorder noflipbyteorder onetile noonetile blackout xinerama noxinerama xrandr noxrandr xrandr_mode padgeom quiet q noquiet modtweak nomodtweak xkb noxkb skip_keycodes add_keysyms noadd_keysyms clear_mods noclear_mods clear_keys noclear_keys remap repeat norepeat bell nobell sel nosel primary noprimary cursorshape nocursorshape cursorpos nocursorpos cursor show_cursor noshow_cursor nocursor xfixes noxfixes xwarp xwarppointer noxwarp noxwarppointer buttonmap dragging nodragging pointer_mode input_skip debug_pointer dp nodebug_pointer nodp debug_keyboard dk nodebug_keyboard nodk deferupdate defer wait nap nonap sb screen_blank fs gaps grow fuzz progressive noremote aro= display vncdisplay desktopname desktop auth rootshift scale_str scaled_x scaled_y scale_numer scale_denom scale_fac scaling_noblend scaling_nomult4 scaling_pad scaling_interpolate inetd safer unsafe passwdfile using_shm logfile o rc norc h help V version lastmod bg nofb sigpipe threads clients client_count pid ext_xtest ext_xkb ext_xshm ext_xinerama ext_overlay ext_xfixes ext_xdamage ext_xrandr rootwin num_buttons button_mask mouse_x mouse_y bpp depth indexed_color dpy_x dpy_y rfbport rfbwait rfbauth passwd alwaysshared dontdisconnect httpdir enablehttpproxy -sync By default -remote commands are run asynchronously, that is, the request is posted and the program immediately exits. Use -sync to have the program wait for an acknowledgement from the x11vnc server that command was processed. On the other hand -query requests are always processed synchronously because they have wait for the result. Also note that if both -remote and -query requests are supplied on the command line, the -remote is processed first (synchronously: no need for -sync), and then the -query request is processed in the normal way. This allows for a reliable way to see if the -remote command was processed by querying for any new settings. Note however that there is timeout of a few seconds so if the x11vnc takes longer than that to process the requests the requestor will think that a failure has taken place. -noremote Do not process any remote control commands or queries. A note about security wrt remote control commands. If someone can connect to the X display and change the property VNC_CONNECT, then they can remotely control x11vnc. Normally access to the X display is protected. Note that if they can modify VNC_CONNECT, they could also run their own x11vnc and have complete control of the desktop. If the "-connect /path/to/file" channel is being used, obviously anyone who can write to /path/to/file can remotely control x11vnc. So be sure to protect the X display and that file's write permissions. -unsafe If x11vnc is running as root (e.g. inetd or Xsetup for a display manager) a few remote commands are disabled (currently: id:pick, accept:, and gone:) because they are associated with running external programs. If you specify -unsafe, then these remote control commands are allowed when running as root. When running as non-root all commands are allowed. See -safer below. -safer Even if not running as root, disable the above unsafe remote control commands. -deny_all For use with -remote nodeny: start out denying all incoming clients until "-remote nodeny" is used to let them in. These options are passed to libvncserver: -rfbport port TCP port for RFB protocol -rfbwait time max time in ms to wait for RFB client -rfbauth passwd-file use authentication on RFB protocol (use 'storepasswd' to create a password file) -passwd plain-password use authentication (use plain-password as password, USE AT YOUR RISK) -deferupdate time time in ms to defer updates (default 40) -desktop name VNC desktop name (default "LibVNCServer") -alwaysshared always treat new clients as shared -nevershared never treat new clients as shared -dontdisconnect don't disconnect existing clients when a new non-shared connection comes in (refuse new connection instead) -httpdir dir-path enable http server using dir-path home -httpport portnum use portnum for http connection -enablehttpproxy enable http proxy support -progressive height enable progressive updating for slow links Pretty wild huh? [1]Contact me if you have any questions or problems. Personally, I use: x11vnc -rfbauth $HOME/.vnc/passwd -nap -flashcmap -cursor X -norepeat -add_keys yms (the -flashcmap only matters on old 8-bit X displays) References 1. mailto:xvml@karlrunge.com