HomeAboutMailing ListList Chatter /0/0 34.228.30.69

SOLVED: Why I can't 'ssh -fCXY /usr/bin/gnome-terminal' ? Grrrrr....

2018-12-05 by: kitepilot@kitepilot.com
From: kitepilot@kitepilot.com
------------------------------------------------------
I found the solution:
ssh -fCXY etroconis@ETROCONIS02-DT '/usr/bin/gnome-terminal 
 --disable-factory' 

Works...
Go figure...   :(
ET 


kitepilot@kitepilot.com writes: 

> The STUPID terminal is starting in the local box, so now the question is:
> why is it ignoring the DISPLAY variable?
> Grrrrrrr...
> ET  
> 
> kitepilot@kitepilot.com writes:  
> 
>> Hello all:
>> This is so freaking odd that I'll explain it with a series of commands:   
>> 
>> There are 2 computers involved, linuxbox (Mint Tara) and wincrap (Win 7).  
>> 
>> In linuxbox I type:   
>> 
>> etroconis@ETROCONIS02-DT:~$
>> ssh -fCXY localhost xclock   
>> 
>> And I get my banana (xclock starts)   
>> 
>> Then   
>> 
>> etroconis@ETROCONIS02-DT:~$
>> ssh -fCXY localhost gnome-terminal   
>> 
>> and (AMAZINGLY!!!!) I get another terminal window.
>> So far so good.
>> Linux is B-O-R-I-N-G...   
>> 
>> Now we move to a cygwin shell in wincrap and I type:   
>> 
>> etroconis@ETROCONIS-DT:~ $
>> DISPLAY=:0.0 ssh -fCXY etroconis@ETROCONIS02-DT xclock   
>> 
>> And (voila), predictably and reliably I again get my banana-clock,
>> so FINAL ACT:   
>> 
>> etroconis@ETROCONIS-DT:~ $
>> DISPLAY=:0.0 ssh -fCXY etroconis@ETROCONIS02-DT /usr/bin/gnome-terminal   
>> 
>> But nope...    8(
>> My terminal never starts in the WinCrap box!   
>> 
>> So, In summary:
>> Please look at the 2 lines together (can you tell the difference?)
>> DISPLAY=:0.0 ssh -fCXY etroconis@ETROCONIS02-DT xclock
>> DISPLAY=:0.0 ssh -fCXY etroconis@ETROCONIS02-DT /usr/bin/gnome-terminal   
>> 
>> I know that things work locally.
>> I know that clock works remotely.
>> I know that gnome-terminal DOES NOT works remotely.
>> Does anybody have any idea why?   :(
>> Thanks for reading my rant...
>> ET
>> 

=============================================================== From: Billy ------------------------------------------------------ Sounds like the Xserver in Windows doesn’t support all the X11 features or extensions that gnome-terminal assumes are there. --b

=============================================================== From: kitepilot@kitepilot.com ------------------------------------------------------ Nope, that's not the problem, it's a gnome booboo. Once I added the switch, it worked perfectly. The story is longer than that, feel free to research it. Best! ET

=============================================================== From: Billy ------------------------------------------------------ So according to this[1], gnome-terminal appears to be creating a dbus instan= ce to handle multiple gnome-terminal windows from the same instance. That also means it won=E2=80=99t connect to multiple X11 endpoints if it=E2=80= =99s using the dbus to launch - the first terminal instance will just use it= s X11 endpoint for all new terminal requests. Also the disable-factory option was undocumented and removed at some point, b= ut Ubuntu 16.04 added a python wrapper[2] to mimic the old option to basical= ly create its own dbus instance with a randomized name to launch its own gno= me terminal - thus allowing multiple terminal instances to connect to multip= le X11 endpoints. 1: https://askubuntu.com/questions/729034/gnome-terminal-disable-factory-sup= port-problem 2: https://gist.github.com/ecatmur/00893506a23e828c6688 --b olution: ctory' Works... s: : There are 2 computers involved, linuxbox (Mint Tara) and wincrap (Win 7)= . In linuxbox I type: etroconis@ETROCONIS02-DT:~$ n etroconis@ETROCONIS02-DT:~$ terminal window. I type: etroconis@ETROCONIS-DT:~ $ , predictably and reliably I again get my banana-clock, nal But nope... 8( nal I know that things work locally.

=============================================================== From: kitepilot@kitepilot.com ------------------------------------------------------ Is it my imagination or is it for real that some people is hellbent in 'windownizing' Linux? No amount of anything (money included) would force me to run systemd in an Internet-facing server. And this is getting more difficult by the minute, maybe it is time to dust off my LFS scripts... :( ET Billy writes:

=============================================================== From: Billy ------------------------------------------------------ Disclaimer: I work for Red Hat. Fedora, CentOs, RHEL 7&8 all use systemd. * =E2=80=9CWhat I find puzzling about all this is that even though everyone a= dmits that sysvinit needed replacing and many people dislike systemd, the di= stributions keep adopting it. Only a few distributions, including Slackware,= Gentoo, PCLinuxOS, and Chrome OS, haven't adopted it.=E2=80=9D [1] * The Biggest Myths of systemd [2] by Lennart Poettering. * Distros using it by default: Arch Linux, CoreOs (pre and post acquisition)= , CentOs, Debian, Fedora, Mageia, Mint, OpenSuse, RHEL, Solus, SLES, Ubuntu.= [3] * My notes and opinion: I have customers using it in production that are in the Fortune 500. They us= e it on thousands of systems. One of the customers I=E2=80=99ve worked with (= not mine) manages 60,000 RHEL and Suse servers - (all running systemd). I have a customer managing thousands (2000+) of remote locations with system= d enabled VMs at each site (RHEL7). I have hospitals and healthcare orgs running systemd enabled systems that mo= nitor patient information: from diabetes, to cancer, to paying claims, and m= anaging care schedules. One (global healthcare) just rolled out a new soluti= on, built on our technology (including RHEL7) that has been verified to have= saved the lives of 7,500 patients in the past 6 months. I have manufactures using systemd enabled systems to build cars, elevators, a= nd handle repair schedules, billing, and customer facing web sites.=20 So, I say all this, because there ARE plenty of organizations out there that= are using Linux in big, expensive, and massively scalable ways - and system= d is NOT an issue, nor even a concern for them. At the end of the day, does the platform enable them to make more widgets, r= educe operational costs, and improve customer or patient experience? Aka: do= es it enable the business to be the best business it can be? The answer is yes. As Linus said in the article [1] - the rest is just detai= ls. 1: https://www.zdnet.com/article/linus-torvalds-and-others-on-linuxs-systemd= / 2: http://0pointer.de/blog/projects/the-biggest-myths.html 3: https://en.m.wikipedia.org/wiki/Systemd --b ndownizing' Linux? Internet-facing server. off my LFS scripts... :( tance to handle multiple gnome-terminal windows from the same instance. That= also means it won=E2=80=99t connect to multiple X11 endpoints if it=E2=80=99= s using the dbus to launch - the first terminal instance will just use its X= 11 endpoint for all new terminal requests. Also the disable-factory option w= as undocumented and removed at some point, but Ubuntu 16.04 added a python w= rapper[2] to mimic the old option to basically create its own dbus instance w= ith a randomized name to launch its own gnome terminal - thus allowing multi= ple terminal instances to connect to multiple X11 endpoints. 1: https://asku= buntu.com/questions/729034/gnome-terminal-disable-factory-support-problem 8, at 4:45 PM, kitepilot@kitepilot.com wrote:=20 e solution: factory' Works... is: ds: There are 2 computers involved, linuxbox (Mint Tara) and wincrap (Win 7= ). In linuxbox I type: etroconis@ETROCONIS02-DT:~$ hen etroconis@ETROCONIS02-DT:~$ er terminal window. nd I type: etroconis@ETROCONIS-DT:~ $ a), predictably and reliably I again get my banana-clock, minal But nope... 8( minal I know that things work locally.

=============================================================== From: Lynn Dixon ------------------------------------------------------ I am also a Red Hatter (Work for Red Hat). At first I didn't like systemd when I was first exposed to it (which was before I was a Red Hatter), but the more I used it the more I actually liked it. The verbosity in why a service fails to start for example is pretty awesome. I can quickly diagnose why httpd or postgresql didn't start correctly at the micro level. I have recently written a systemd service file in one of the projects I started on my github: https://github.com/lynndixon/LiveSectional I was pretty shocked at how easy it was to create a powerful systemd service that would launch, monitor and restart my sort-of-janky python daemon service for me automatically. Granted this is my first foray into Python and my code is pretty rough, but having systemd so easily monitor it and restart the daemon when my crappy python code breaks was a pleasant bit of work. Now being a Red Hatter, I also work with very very large customers that love and use systemd / RHEL daily. I'd have to +1 everything Billy said. Folks use it and like it every day in the enterprise. It scales superbly, and does what it needs to do. ne e , with (not w d , : / n t . =E2=80=99s using as . port-problem

=============================================================== From: Ed King ------------------------------------------------------ ology (including RHEL7) that has been verified to have saved the lives of 7= ,500 patients in the past 6 months. Saved it or prolonged it? I thought we all had to exit(0) or exit(1) eve= ntually X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KQ2h1Z2FsdWcg bWFpbGluZyBsaXN0CkNodWdhbHVnQGNodWdhbHVnLm9yZwpodHRwOi8vY2h1Z2FsdWcub3JnL2Nn aS1iaW4vbWFpbG1hbi9saXN0aW5mby9jaHVnYWx1Zwo=

=============================================================== From: Stephen Kraus ------------------------------------------------------ They used a Turboencapulator to implement the solution, I'm sure of it On Fri, Dec 7, 2018, 1:29 PM Ed King >One (global healthcare) just rolled out a new solution, built on our

=============================================================== From: Billy ------------------------------------------------------ Let=E2=80=99s just say it prevented an exception or segfault, and allows the= ir life thread to complete whatever work it needs to accomplish. --b nology (including RHEL7) that has been verified to have saved the lives of 7= ,500 patients in the past 6 months. entually