x2go-dev March 2017

x2go-dev@lists.x2go.org

20 participants
139 discussions

Bug#1154: x2goagent nightly build for Fedora 25 is not signed
by Mike DePaulo 19 Apr '17

19 Apr '17

Package: nx-libs Version: 3.5.0.33 When I attempt to update to X2Go nightly builds on Fedora via these instructions: http://wiki.x2go.org/doku.php/wiki:repositories:fedora (Last updated 2016-11-01) I receive this error: Error: Package x2goagent-3.5.0.33-0.0x2go1.0.git20170309.756.heuler.fc25.x86_64.rpm is not signed I worked around this by passing --nogpgcheck: sudo dnf update --nogpgcheck

3 2

Sponsorship of X2Go macOS Client Development and of X2Go Linux on POWER Development
by Stefan Baur 09 Apr '17

09 Apr '17

Hi everyone, as previously announced, CFD.direct has decided to sponsor our macOS Client Development, and XCLOUD.me, has jumped right in and offered us a special rebate, so we can run two of their smallest instances for the price of one medium instance. This means that will be able to offer automated X2GoClient for macOS builds both for macOS 10.10 (these should work on 10.9 and 10.8 as well) as well as for macOS 10.12. The machines have been provisioned today, and Mihai will be setting them up over the course of the next few days. Once he's done with that, we'll move on to supporting the POWER platform directly again (at present, some distributions grab our source and turn it into packages for their POWER flavor, but there's no official X2Go repository where you could get stable or nightly builds directly from us). IBM and the Oregon State University's Open Source Lab have teamed up and are providing us with four POWER8 virtual machines (with the option to add more, should the need arise), manageable via an OpenStack WebGUI. Unlike the previous POWER8 virtual machine that we had, these new servers are going to be available for us for the foreseeable future. We intend to run at least two (one for Big-Endian, the other for Little-Endian) of those machines as Jenkins Build Slaves, just as the virtual macOS machines, so we will soon be able to provide automated builds for POWER. The other two we currently intend to run as demo servers and for testing and bugsquashing, though we might repurpose them as Jenkins Build Slaves if it turns out that we need one per Endianess *and* packaging system (DEB vs. RPM). There's one bug that's cropped up on POWER8 LE only, and we're already curious to learn what's causing it, so we're glad we finally have access to a test environment again. Here is a quick overview over the sponsors mentioned above: CFD Direct <https://cfd.direct> are the primary developers of OpenFOAM <https://openfoam.org>, the free, open source software for computational fluid dynamics. XCLOUD.me <https://xcloud.me/> is a hosting company that specializes in providing virtual macOS cloud instances - all 100% legal, as the virtual machines run on top of real Apple hardware. IBM are the original developers of the POWER systems used in midrange computing. You can learn more about the Linux on POWER ecosystem here: <http://www-03.ibm.com/systems/power/software/linux/> Oregon State University's Open Source Lab (OSU/OSL) is a nonprofit organization working for the advancement of open source technologies. They offer various means of support to open source projects, one of them being free hosting on both Intel and POWER servers. To find out more about them, please visit their web site at <http://osuosl.org/> Kind Regards, Stefan Baur X2Go Project/Community Manager -- BAUR-ITCS UG (haftungsbeschränkt) Geschäftsführer: Stefan Baur Eichenäckerweg 10, 89081 Ulm | Registergericht Ulm, HRB 724364 Fon/Fax 0731 40 34 66-36/-35 | USt-IdNr.: DE268653243

2 3

X2Go client to search for server?
by John Cobo 01 Apr '17

01 Apr '17

Hello, I am considering X2Go for a project which involves non-technical people using X2Go to connect to a Raspberry Pi which does not have a screen and so the user will not know the Pi's IP address. I could set an SSH port on the Pi to an obscure number such as 2432 or something. Would it be feasible for the X2Go clients have a new option to scan a range of IPs (eg. 192.168.0.1 - 192.168.0.255) for a given port (eg. 2432) on which to connect? Such a feature could solve the generic problem of how to connect for the first time to something new on your local network. Thank you, John Cobo

3 8

Bug#1069: Actual bug is in Tk
by Stefan Baur 31 Mar '17

31 Mar '17

Hi, after some more research, it turns out that the actual bug is in Tk. We'll try to locate the code in question and report it upstream, possibly even sending a patch along. Current ETA (subject to change): End of April, 2017 -Stefan -- BAUR-ITCS UG (haftungsbeschränkt) Geschäftsführer: Stefan Baur Eichenäckerweg 10, 89081 Ulm | Registergericht Ulm, HRB 724364 Fon/Fax 0731 40 34 66-36/-35 | USt-IdNr.: DE268653243

1 0

Bug#1109: Print all In new window Long delays (i.e. several minutes) waiting for redraw ONLY in seamless window mode
by Aki Ketolainen 31 Mar '17

31 Mar '17

Hi, I noticed yesterday that disabling "show window contents while moving" in Mate desktop and XFCE fix this behaviour. I haven't found the setting to do this in Unity, though. Also KDE, LXDE and Windows on the client side don't show this behaviour. @sunweaver on #x2go on freenode commented that "21:55 <@sunweaver> I guess a sustainabe fix is grabbing the window content client side and keep that static pixmap until movement is over." Best regards, Aki

1 0

Bug#1010: Slow repaint on window move (single application)
by Aki Ketolainen 31 Mar '17

31 Mar '17

1 0

Re: [X2Go-Dev] [X2Go-Commits] [x2goclient] 01/01: Don't override PATH for the actual session or application command. Fixes: #1100
by Mihai Moldovan 30 Mar '17

30 Mar '17

On 26.03.2017 06:11 PM, git-admin(a)x2go.org wrote: > This is an automated email from the git hooks/post-receive script. > > x2go pushed a commit to branch bugfix/1100 > in repository x2goclient. > > commit a7ed6868825c111f8d0fa4a64aa82115b8dab039 > Author: Mike DePaulo <mikedep333(a)gmail.com> > Date: Sun Mar 26 12:09:11 2017 -0400 > > Don't override PATH for the actual session or application command. Fixes: #1100 > --- > debian/changelog | 3 ++ > src/onmainwindow.cpp | 67 ++++++++++++++++++++++++++++----------------- > src/onmainwindow.h | 1 + > src/sshmasterconnection.cpp | 4 +-- > src/sshmasterconnection.h | 2 +- > src/sshprocess.cpp | 15 ++++++++-- > src/sshprocess.h | 2 +- > 7 files changed, 62 insertions(+), 32 deletions(-) > > diff --git a/debian/changelog b/debian/changelog > index 42c1e52..8e71aeb 100644 > --- a/debian/changelog > +++ b/debian/changelog > @@ -180,6 +180,9 @@ x2goclient (4.1.0.1-0x2go1) UNRELEASED; urgency=medium > default because the installation dir is not writeable by > users) > + CVE-2017-6542 was fixed > + - Don't override PATH for the actual session or application > + command. > + Fixes: #1100 > > [ Seth Galitzer ] > * New upstream version (4.1.0.1): > diff --git a/src/onmainwindow.cpp b/src/onmainwindow.cpp > index 5dd3906..0b15649 100644 > --- a/src/onmainwindow.cpp > +++ b/src/onmainwindow.cpp > @@ -6291,7 +6291,7 @@ void ONMainWindow::slotProxyStderr() > { > xmodExecuted=true; > QTimer::singleShot ( > - 2000, this, > + 4000, this, Why are you changing this? It's orthogonal to all the other changes. > SLOT ( slotExecXmodmap() ) ); > } > } > @@ -6700,6 +6700,39 @@ void ONMainWindow::slotAppDialog() > > void ONMainWindow::runCommand() > { > + > + if ( runRemoteCommand ) > + { > + /* 1st override PATH and determine the base path to x2goruncommand. Whitespace errors, you're mixing tabs and spaces here. > + * Then in SlotRunCommand, call x2goruncommand without overriding PATH. > + * This ensures that the PATH is never overriden with for the actual "overridden", remove "with" or specify. > + * user session. > + * Fixes: #1100 > + */ > + sshConnection->executeCommand ( "x2gobasepath", this, > + SLOT ( SlotRunCommand ( bool, > + QString, > + int )), true); > + } > +#ifdef Q_WS_HILDON > + //wait 5 seconds and execute xkbcomp > + QTimer::singleShot ( 5000, this, SLOT ( slotExecXmodmap() ) ); > +#endif We'll probably execute slotExecXmodmap a bit sooner than before with that change (since you're not doing it in SlotRunCommand - but that wouldn't work for XDMCP and Shadow sessions, since they don't set runRemoteCommand to true.) It's fine, I guess. I just noticed that the code in question is broken anyway - it wouldn't even compile on HILDON anymore, so let's ignore that. Otherwise looks good. Mihai

2 1

Bug#1167: Windows 7 runtime error - R6025 - pure virtual function, kills x2go client after Windows screen lock/unlock
by Petronic, Mark 29 Mar '17

29 Mar '17

Package: x2goclient Version: 4.1.0.0 Steps to reproduce: 1. Fire up x2go client 2. Establish a session 3. Session starts fine and all is well 3. Stop working and walk away and, due to IT policy, Windows screen will lock after 15 minutes of inactivity 4. Ctrl-Alt-Delete and log back into Windows 5. Find a popup error dialog indicating vcxsrv.exe crashed. The runtime error Windows popup message dialog says: ========================================================= Microsoft Visual C++ Runtime Library Program: C:\Users\mpetronic\<path-to-my-install>\VcXsrv\vcxsrv.exe R6025 - pure virtual function call ========================================================= I can also force the crash if, after I establish a working x2go session, I Ctrl-Alt-Delete and lock my desktop manually. Same crash occurs when I log back in. To recover, I have to close the x2go client and restart it then I can restart my session. On the server side, my session is restarted just fine. Everything I had open there is restored. So, this is really annoying in that, every time I walk away from my PC for a short time, I know I have to reestablish all my sessions due to this crash. I have further experimented and found this crash ONLY occurs when I run in -rootless configuration. I use that so that Alt-Tab works properly in my X session otherwise, when I Alt-Tab, I bounce back to windows in Windows. I basically use these settings set in the "X.Org Server Settings" dialog per (http://www.terheyden.com/blog/?p=202) -rootless -notrayicon -clipboard -keyhook Another tidbit of information is this crash only started to happen a few months ago but I made no changes to my x2go client installation or configuration. It started after a Windows update. Background: To access our remote systems at work, we have to use VMWare's VDI (Virtual Desktop Infrastructure). So, I have to first log into my corporate desktop, then start a VDI session using RSA key login, which gives me another Windows VDI desktop. On the VDI Windows instance, I have x2go installed in the portable way because I don't have admin access these to install it by running the setup program. Some months back, the VDI admins installed some Windows update per normal security patching. After that, this crash started happening. There happens to be a old VDI configuration that apparently has NOT been patched. If I log onto that instance, the crash does not occur. This crash happens with your latest version 4.1.0.0 and two previous versions so it appears to be related to something common to all versions. Since we still have this old (working) VDI configuration (until some realizes it and updates it) and new (broken) one, maybe there is a way I can help by getting you some information as to what is different between the two related to the Windows run time libraries installed? I am glad to help in any way to get you information to help in debugging and fixing this issue. Please let me know what I can do to assist. Thanks! [cid:image001.png@01D2A87C.0E8D7B90]

1 0

Bug#1166: x2goclient terminates with "WARNING: failed to kill process group '<pid>': No such process
by Seth Galitzer 28 Mar '17

28 Mar '17

Package: x2goclient Version: 4.1.0.0 and 4.1.0.1 When exiting x2goclient, after a pause of about 2 seconds, the following message is displayed to stderr: WARNING: unable to send SIGTERM to process group '<pid>': No such process About 10 more seconds later, the following message is displayed to stderr: WARNING: failed to kill process group '<pid>': No such process In some cases, the first message is not displayed, but the second one always is after about 10-12 seconds. The pid is the same in both message. This seems to only affect 4.1.x clients. 4.0.x clients do not seem to have this behavior. I have confirmed these results on the following platforms and builds: Ubuntu 16.04 (xenial): x2goclient 4.1.0.0 (4.1.0.0-0~1536~ubuntu16.04.1), installed from Launchpad PPA Ubuntu 16.04 (xenial): x2goclient 4.1.0.1, compiled from source Ubuntu 15.04 (vivid): x2goclient 4.1.0.0 (4.1.0.0-0~1536~ubuntu15.04.1), installed from Launchpad PPA Ubuntu 15.04 (vivid): x2goclient 4.1.0.1, compiled from source Debian 8.7 (jessie) 64-bit: x2goclient 4.1.0.1, compiled from source Debian 8.7 (jessie) 32-bit: x2goclient 4.1.0.1, compiled from source The following builds and platforms do not appear to be affected: Ubuntu 16.04 (xenial): x2goclient 4.0.5.1 (4.0.5.1-1), installed from official Ubuntu repos Ubuntu 15.04 (vivid): x2goclient 4.0.3.1 (4.0.3.1-4), installed from official Ubuntu repos Debian 8.7 (jessie) 64-bit: x2goclient 4.0.3.1 (4.0.3.1-4), installed from official debian repos Debian 8.7 (jessie) 32-bit: x2goclient 4.0.3.1 (4.0.3.1-4), installed from official debian repos Seth -- Seth Galitzer Systems Coordinator Computing and Information Sciences Kansas State University http://www.cis.ksu.edu/~sgsax sgsax(a)ksu.edu 785-532-7790

3 2

Bug#1085: 4.1 upgrade has not fixed the crashing
by Troels Arvin 27 Mar '17

27 Mar '17

Upgrading to X2Go 4.1.0.0: Unfortunately, the crashes still occur.

1 0

← Newer
1
...
7
8
9
10
11
12
13
14
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

x2go-dev March 2017