Re: [X2Go-Dev] X2go, printing, Vserver, PostgreSQL - Progress....
On Sun, 2012-09-30 at 10:35 -0400, senrab...@aol.com wrote: Hi All: Another quick update - we think enabling fuse in the vserver guest is part of the problem, though the vserver folks suggest this may be a security/stability problem. In any event, that fixed the broken spool link, and we think we're on to cups admin how to issues that may be solved outside of this forum. More to follow... snip Newer kernels may break out the capability required to make FUSE work from the admin capability but I've not investigated that yet. If you allowed the admin capability in your vserver guest, you shot your security to bits. If I recall correctly, the capability limitation was not in mounting FUSE drives but only in unmounting them, strangely. That's why we moved the x2gocleansessions script to the VServer host - not to mention that it means we can run one process for many hundreds of servers rather than one each firing every five seconds. We do have this working without opening the admin capabilities but I do not remember the details off the top of my head and we are using an old and heavily adapted version. Good luck with it - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
[X2Go-Dev] Proxy versus broker?
Hello, all. I was delighted to see the pending release of proxy capability. Is this at all related to the project undertaken quite a while ago to create a broker, i.e., to have a single publicly exposed site to which users would connect and then be directed to their appropriate X2Go Server? We currently struggle with assigning unique X2Go servers (and we use one per user via VServer) unique ports to avoid consuming countless public IP addresses. It would be so much easier of all could connect to a single device and have that device broker the connection to their internal X2Go Server. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Bug#34: SSH_OPTIONS_FD
On Tue, 2012-09-25 at 11:13 +0200, Oleksandr Shneyder wrote: Am 25.09.2012 11:01, schrieb Moritz Struebe: On 2012-09-25 10:47, Oleksandr Shneyder wrote: Sure, it is a fail of system administrator, if he allow such unecrypted authentication over Internet. But I don't even give them a possibility to make such mistake... Sorry, here should be I don't want to give instead of I don't give I don't really get your point. The credentials are used by the browser anyway - because otherwise there would be no need for a proxy. I don't think it's our job to disable features because of incompetent system administrators. After all proxy authentication is normally used within LANs. I don't want to disable any features. I only say, it is nice to have a possibility to send authentication data to server encrypted. In LAN it is not a such big problem to send it in clear text. But in case of SSH-Proxy it is an Internet connection. And I want, that every one, who use this feature with X2Go know, that sending unecrypted data over Internet is not safe. And that should not be the same authentication data as used on other servers. snip I very much agree with Alex here. Although we can absolve ourselves of the responsibility, it is wiser to do as much as we can to prevent both admins and users from shooting themselves the stupid things they may do. For example, it is not just a matter of a sloppy admin not realizing they should use a separate authentication domain for the proxy; even if they do, we have the social engineering problem of users using the same password for the proxy as for anything else. Once one intercepts that password, a cracker will try it everywhere they can for that user. Thus, I would strongly advocate all authentication even to the proxy be protected by encryption. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] X2go, printing, Vserver, PostgreSQL - PS
On Sun, 2012-09-23 at 09:54 -0400, senrab...@aol.com wrote: Hi All: This morning, per yesterday's email re: our trouble printing using a Vserver guest with postgreSQL and a Vserver host doubling as a cups server, we tried an experiment that didn't work but thought we'd add it in. -We increased the Vserver config fstab (not the fstab in the Vserver's /etc but the one in the Vserver config folder) to 1024M. -We started x2go and noted the session name -By hand, we created a folder with the session name in /tmp (i.e., tmp/.x2go-user1/spool/C-user1-50-1348405390_stDGNOME_dp24) on the Vserver guest/x2go server. -We erased the broken spool link, and created a new one that points to the new folder in /tmp again on the Vserver guest/x2go server. At that point, the link was active (not broken) but nothing changed in our inablity to print. I.e., we start a print job on the virtual x2go printer, all seems to go fine, but no joy in terms of evince starting on the client or getting a page to print. Still though, the print jobs are showing up in /var/spool/cups on the cups server, in the cups interface (http://localhost:631), the page_log in /var/log/cups on the cups server, and we don't get any errors in error_log in /var/log/cups on the cups server. That has been the case before. We think that fact that we can connect via x2go shows that the postgreSQL setup is working, that the print jobs are showing up on the cups server suggests that the printing ssh connection works (though we're less sure about this). Not sure what to make of thisbut hope it was worth adding. snip How is the CUPS server establishing its SSH connection to the x2goserver? - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] X2go, printing, Vserver, PostgreSQL
On Sat, 2012-09-22 at 11:17 -0400, senrab...@aol.com wrote: Hi All: We're trying to get printing to work using X2go and Vserver. Per Mike's suggestion, we've shifted to trying the Postgres approach (http://www.x2go.org/doku.php/wiki:advanced:multi-node:x2goserver-pgsql and http://www.x2gorg/doku.php/wiki:advanced:multi-node:x2goserver-printing) but are still stuck. This is long, but are hoping will help... Some suspects are DNS, permissions (thanks John Sullivan for his suggestion), and ssh between the x2go server on the vserver and the cups server on the host (see below...Set up SSH Keys) Right now, when we try to print, superficially things seem to go as they should (e.g., entries for successful printing get entered in the cups logs) but nothing happens either in terms of popping a pdf and getting Show this diaglog before starting printing on the client or getting a page to print. FYI - the symptoms are pretty much the same as when we were stuck with single node x2go printing. TERMS: We think of things in the following terms, and hope this is useful for clarity: Host - computer that runs Vserver platform and Cups server. It is on this that we have set up the virtual x2go printer, shared it, and generated the ssh key for printing from x2go servers Guest - Vserver instances where x2goserver and the PostgreSQL backend are located and x2goserver-printing is installed, to which a client connects Client - netbook from which we connect to x2go on the Vserver Guest. SETUP 1) On Vserver Guest (x2go server): - Installed and set up PostgreSQL; added the user to postgres (user1) to be used when connecting from the x2go client to the x2go server and PostgreSQL database (x2godgadmin --adduser user1). - Installed x2go-server-printing NOTE: we got the following messages when setting up PostgreSQL and because we can connect (and sqlpass gets generated in user's .x2go on the x2go server) assume they don't matter, but have included them: root@vserver1:/usr/lib# x2godbadmin --createdb NOTICE: database x2go_sessions does not exist, skipping NOTICE: role x2godbuser does not exist, skipping NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index sessions_pkey for table sessions NOTICE: CREATE TABLE / UNIQUE will create implicit index sessions_display_key for table sessions NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index messages_pkey for table messages NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index used_ports_pkey for table used_ports NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index mounts_pkey for table mounts root@vserver12:/usr/lib# x2godbadmin --adduser user1 NOTICE: role x2gouser_user1 does not exist, skipping create DB user x2gouser_user1 2) On Vserver Host (Cups Server): - Installed cups-x2go - Used http://localhost:631; and added the virtual x2go printer and shared it (we couldn't see it on the client unless we shared it) - Used Visudo to add x2goprint ALL=(ALL) NOPASSWD: /usr/bin/x2goprint - Set up SSH Keys using http://www.x2go.org/doku.php/wiki:advanced:multi-node:x2goserver-printing NOTE: we had trouble getting the keys to copy over (ssh-copy-id /usr/bin/ssh-copy-id: ERROR: No identities found). We cannot ping the Guest from the Host by server name (e.g., vserver1.mydomain.com), but can by ip address. We ended up creating /home/x2goprint/.ssh/authorized_keys and adding the contents of id_dsa-x2goprint.pub by hand. We then issued ssh -i /root/.ssh/id_dsa-x2goprint x2goprint@x192.168.1.112 and got: Oops! This email landed in my SPAM folder so I didn't see it before my previous reply. This definitely looks like a problem in that the CUPS server is not going to be able to scp the file to the x2goserver like this. If you can't manually scp a file from the print server to the X2Go server, the print server won't be able to do it either - at least so I would think. The authenticity of host '192.168.1.112 (192.168.1.112)' can't be established. RSA key fingerprint is dd:04:0f:56:5f:23:a8:71:e6:d8:aa:64:4c:91:16:0d. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added '192.168.1.112' (RSA) to the list of known hosts. Permission denied (publickey). Is this Permission denied (publickey). the cause of our problems? It appears to have added an RSA key where we think the keys created where DSA... 3) Tested the x2go connection from the client. We could connect. Then tried to print: we can see the virtual x2go printer but cannot get any of the desired outcomes to occur (printing, get pdf to open using evince parameter). 4) Observations On the Vserver host (cups server) a) the print job files show up in /var/spool/cups b) in http://localhost:631 the print jobs show as completed c) there are no errors in /var/log/cups/error_log and the page_log and access_log seem to restister the
Re: [X2Go-Dev] Printing, x2go, Vservers
On Sun, 2012-09-16 at 13:59 -0400, senrab...@aol.com wrote: Hi All: snip 3) John's Response Does the CUPS server have rights to write to the client's spool directory? Good luck - John Hi John - how are you? Good question...spent some time trying to check by comparing server where we connect directly to the x2go server vs. x2go running on a Vserver guest including on both: /tmp/.x2go-user1/spool/C-user1-50-1347811312_stDGNOME_dp24 (see below) /var/spool/cups Even if we (assuming no mistakes) make the permissions comparable on both the x2go to x2go server and the x2go to x2go server on Vserver guest on these folders, no joy. This may be the right track, thougha couple of observations/questions: a) On the box where we connect directly the x2go server (no vserver involved), in the .x2go folder for the user there is a healthy spool link - on the vserver example, this link exists but is broken (they both point at /tmp); Bertl at Vserver says this is OK. b) When we try to print, on the working x2goclient to x2goserver (again, no vserver), the /tmp/.x2go-user1/spool/C-user1-50-1347811312_stDGNOME_dp24 folder gets created along with the pdf for the print job. NOTE: this does not happen on the x2go client to x2go server on Vserver guest. I.e., no comparable folder is created. So this folder which we think gets generated on the fly, only gets generated on the x2go to x2go server (not the instance where we connect to the x2go server on the vserver guest). snip If I recall correctly, the x2go printer driver transfers the file to the X2Go Server via scp. Can the CUPS server establish an scp session to drop off the file? - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Printing, x2go, Vservers
On Sat, 2012-09-15 at 13:49 -0400, senrab...@aol.com wrote: Hi All: We are trying to get x2go printing to work on a Debian Squeeze Vserver guest. 1) First, tested without Vserver - set up an x2go virtual printer on a box running x2go server; connected via an x2go client and could print (used steps from http://www.x2go.org/doku.php/wiki:components:printing). 2) On another box running Vserver, set up Vserver Guest and replicated our steps to set up x2go printing including a virtual x2go printer - used http://linux-vserver.org/Problematic_Programs#Cups_print_server which entails as we understand it disabling cups listening to loopback 127.0.0.1 on the host so it doesn't conflict with the Vserver, We also adding entries to ccapabilities (SECURE_MOUNT, SECURE_REMOUNT,BINARY_MOUNT) based on some other posts. When we connect via X2go from client netbook to Vserver guest, and try to print, we can see the x2go virtual printer and printing appears to happen, but fails (both trying settings to print directly to a printer and the pdf pop approach using evince). We're not sure how to troubleshoot this...we get an error in the Vserver cups error log /var/log/cups/error_log of Unable to open listen socket for address ::1:631 - Address family not supported by protocol.) that we think is generated by the printing attempt.. This would seem to suggest some problem with sockets(?)... Suggestions appreciated! Any suggestions appreciated. PS: on the same Vserver guest, from x2go on the netbook client we can print to a network printer on the local Vserver guest/host network. But not the virtual x2go virtual printer. snip Does the CUPS server have rights to write to the client's spool directory? Good luck - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] bug#18: Connecting via SSH proxy fails after a control session got disconnected due to network timeouts
On Tue, 2012-09-11 at 09:35 +0200, Mike Gabriel wrote: Package: python-x2go Version 0.2.0.9 Severity: normal Python X2Go crashes sometimes... 1. session established 2. network fails 3. session receives a network timeout 4. session and connection will get taken down by python-x2go code 5. on re-connect and then session restart/resume the below exception pops up The problem is often seen when SSH proxy connections are used. Not seen yet with direct SSH connections. snip It may not be the same problem but there are problems with the x2goclient 3.99.2.3 (and 3.0.1-4) with direct SSH connections when the network connection drops. One must wait several minutes before one is able to reconnect without problems. I have not taken the time to isolate it yet - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] X2go connection problem to a Vserver
On Sun, 2012-07-29 at 13:15 -0400, senrab...@aol.com wrote: To Whom It May Concern: We are trying to get x2go to work with a Vserver instance but cannot connect. We've done some looking at x2go posts into the error, but we're not sure the responses apply to using Vserver and are still looking for some help (apologies if we should have been able to find this on our own). The error is first: X2go-51 The remote proxy closed the connection while negotiating the session. This may be due to the wrong authentication credentials passed to the server. The window includes an OK button. If we press the OK button, we get to the x2go login screen where you can enter a password...if we enter the password, we get: Error Connection failed channel_open_session failed - Writing packet: error on socket (or connection closed): Broken pipe. On the client, in the sessions file we see: Loop: PANIC! The remote NX Proxy closed the connection. Loop: PANIC! Failure negotiating the session in stage '7'. Loop: PANIC! Wrong version or invalid session authentication cookie. In both the client options file, the cookie looks to be the same as the cookie in the server FYI - we can ping the Vserver instance, and we can ssh into the Vserver successfully, ping 8.8.8.8 from there etc. Also, initially when this wouldn't work, we though maybe we had issues with how we had set up desktops in the Vserver (which is still possible: we're using gnome, gdm) but when we spun up another Vserver and set up VNC, it serves the desktop correctly so at first blush this seems to be a networking/ports/firewall/sshd/routing type of issue. However, we're still having trouble getting a regular X session to run over SSH.but are assuming at the moment (possibly incorrectly) that the X2go errors we're getting are an issue on their own, possibly in addition to running an X session on a Vserver. snip If you are having problems running X over SSH, you probably won't be able to run X2Go - at least that's my guess. Is your /etc/ssh/sshd_config set up to allow X? You may wish to try the following settings: X11Forwarding yes X11DisplayOffset 10 X11UseLocalhost no Good luck - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Fullscreen RDP sessions with X2Go
On Tue, 2012-07-03 at 15:57 +0100, Helmer Teles wrote: snip Can you provide the whole script? It feels like this (or something very much like this) should become part of X2Go upstream. snip Here it goes, it's not even very finished but it works for some of our winblows customers. I'm more on the Linux Desktop KDE 4 Nightmare to kiosk thing ;) I've attached it to you because it will get nasty to get something of inline text. Side note: this script is running in a kind of forgotten by me linux host. Been hard @ working with KDE 4 as DaaS to replace our aging Debian Lenny with kde3 Linux cloud Desktops. snip Interesting. Maybe that's why I have not seen it as we are not using the latest xfreerdp yet. Helmer, as an aside, we sidestepped the KDE4 nightmare by switching to the Trinity desktop. In case you are not familiar with it - www.trinitydesktop.org. It is a maintained fork of KDE3 and has worked well for us. Good luck - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Cut Paste between 2 Ubuntu servers
On Fri, 2012-06-29 at 18:29 +0200, Stefan Baur wrote: Am 29.06.2012 18:22, schrieb Mike Gabriel: So this is really weird then. I use several desktop sessions from my notebook on several servers. I can paste between different sessions on different servers and also between session and the local X display. No matter if sessions run in desktop mode or rootless mode... So I have to scratch my head on this... Could it be that you're using different methods, with one working and the other failing? After all, there's Ctrl-C/Ctrl-V, Ctrl-Ins/Shift-Ins, copying/pasting via right mouse button and context menu, copying/pasting via mouse-select/middle-mouse-button, etc. And IIRC, there are different types of clipboards on Linux, as well. See, for example, this link here: http://mutelight.org/subtleties-of-the-x-clipboard snip Fascinating article. I am a KDE and thus Klipper user. I have noticed that sometimes pasting works perfectly fine whereas other times, I must activate the latest entry in Klipper. I wonder if that has to do with the difference between PRIMARY/CLIPBOARD/SECONDARY? I've noticed particular problems with OpenOffice. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Removal of EsounD support
On Wed, 2012-06-13 at 19:27 +0200, Mihai Moldovan wrote: Hi all, I'm thinking about removing ESD support from x2goclient altogether. During my recent deep changes to the PulseAudio integration for X2goclient on both Windows and Mac OS X, I came across the ESD sound support. On Windows, ESD was supported through the PulseAudio wrapper, while OS X was totally broken anyway. Testing ESD support proves to be difficult, and I'm thinking about removing it completely. Most Linux distributions have already removed esd (and tools) from their repo, or rather replaced it by the PulseAudio esd compat layer. That's especially true for Ubuntu, but also Debian Testing and Unstable (Debian Stable still has esd for x86 and amd64 arches, but this will be dropped in favor of PulseAudio in the upcoming release.) That said, I'd like to know whether users are still using esd or not. I wouldn't feel comfortable removing a feature still in use. Please report, if you find the time. snip Removing it is fine by me - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] occurrence of sluggishness in X2Go
On Wed, 2012-06-06 at 10:10 -0400, John A. Sullivan III wrote: On Fri, 2012-06-01 at 08:32 +0200, Mike Gabriel wrote: Hi, just for the record and to let people know about current work focus... People (esp. John) have reported occurrences of a very sluggish behaviour of various applications when run in X2Go sessions. We developers can now confirm this. Ubuntu users reported this earlier, Debian user can now also experience such a sluggishness. Reproduce on Debian: * add squeeze-backports archive to your system * install iceweasel (esr version) This will drag in a libcairo from squeeze backports that is very probably the cause for making some applications very slow within X2Go sessions. Libcairo is used in all GTK applications and is a layer between GTK and the rendering backend (X11, OpenGL, ...). With some recent version of libcairo the sluggishness got introduced... @John: so it is not Nagle, I guess... Compare: launch gnome-terminal (uses libcairo) on such a system and xterm (uses direct X11 rendering). Now create some output on standard out and then scroll back the shell history... Alex has become aware of this on Linux Day Berlin and he has put it on his list with high priority to find a solution/workaround for that. snip Hmm . . . as I think about it, although I eagerly await this fix, I believe we are looking at two separate issues. We are still running 3.0.1 series on our production servers and clients because of this issue (and some others). However, I personally run the latest client from heuler. I have the problem on both the latest x2goserver and the old 3.0.1-5 x2goserver. It appears to be a client and not a server issue and appeared with the move to libssh. The delay is not inconsequential. We are not talking about a sese of sluggishness with a 250ms delay. We are talking about a 2 to 5 full second delay for the final keystrokes or screens to appear. Everything else until the final bit is lightning fast even on WAN links - John snip I happened to have an old 3.0.1-7 plugin sitting around and tried that. I observed some of the same naglish behavior so it may not be the x2goclient. Unless someone else is experiencing this, I need to investigate our environment more fully before sending the devs on a fruitless search - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] occurrence of sluggishness in X2Go
On Fri, 2012-06-01 at 08:32 +0200, Mike Gabriel wrote: Hi, just for the record and to let people know about current work focus... People (esp. John) have reported occurrences of a very sluggish behaviour of various applications when run in X2Go sessions. We developers can now confirm this. Ubuntu users reported this earlier, Debian user can now also experience such a sluggishness. Reproduce on Debian: * add squeeze-backports archive to your system * install iceweasel (esr version) This will drag in a libcairo from squeeze backports that is very probably the cause for making some applications very slow within X2Go sessions. Libcairo is used in all GTK applications and is a layer between GTK and the rendering backend (X11, OpenGL, ...). With some recent version of libcairo the sluggishness got introduced... @John: so it is not Nagle, I guess... Compare: launch gnome-terminal (uses libcairo) on such a system and xterm (uses direct X11 rendering). Now create some output on standard out and then scroll back the shell history... Alex has become aware of this on Linux Day Berlin and he has put it on his list with high priority to find a solution/workaround for that. snip Hmm . . . as I think about it, although I eagerly await this fix, I believe we are looking at two separate issues. We are still running 3.0.1 series on our production servers and clients because of this issue (and some others). However, I personally run the latest client from heuler. I have the problem on both the latest x2goserver and the old 3.0.1-5 x2goserver. It appears to be a client and not a server issue and appeared with the move to libssh. The delay is not inconsequential. We are not talking about a sese of sluggishness with a 250ms delay. We are talking about a 2 to 5 full second delay for the final keystrokes or screens to appear. Everything else until the final bit is lightning fast even on WAN links - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] PulseAudio
On Thu, 2012-05-31 at 10:17 +0200, Moritz Strübe wrote: Am 31.05.2012 09:19, Stefan Baur schrieb: Do we have any chance to re-code the audio on the server into a lossy, lower-quality version before transmitting it to the client? I mean, we provide quie a few options for image compression, to match different bandwidth limits - would a similar approach be possible for audio as well? As long as video uses the same connection this will not help, as the delay is the problem, not the amount of audio data. The only way to get it working is to encode sound and video together. QoS might help a bit, but only under certain circumstances. I'm afraid at the moment we have to be happy with what we have (which is way better then everything else :-) ). snip Despite the fact that it will not solve this problem, the idea of compressing audio based upon bandwidth is very attractive. I have found video on NX difficult even on a LAN. SPICE handles this much better. As I mentioned several times, as it matures, we may wish to slot it in as an alternative to NX in the X2Go framework - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] First NX/libcairo tests succeeded (was Re: Fwd: Fourth Maintenance Release of the NX 3.5.0 Node and Server packages)
On Tue, 2012-05-22 at 11:18 +0200, Mike Gabriel wrote: Hi all, On Di 22 Mai 2012 10:37:32 CEST Mike Gabriel wrote: snip So my first tests succeeded. Can anyone else confirm the x2goagent from deb http://packages.debian.org/debian wheezy|sid heuler works with GNOME (or iceweasel, icedove)? Please cross-check that session resuming also works as expected, esp. on Debian squeeze systems: deb http://packages.debian.org/debian squeeze heuler snip I can confirm successful session resumption on Squeeze - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] X2Go-Future
On Sun, 2012-05-20 at 14:37 +0200, Mike Gabriel wrote: snip Sadly the expirience suffered when we updated the nx- libs from 3.4 to 3.5: On WAN-connections its still usable but on LAN X2Go looses in speed and latency and responsiveness against VNC. Which my co- workers really hate. Then there must be something wrong with your setup. X2Go works on UMTS (even GPRS) networks as well as on local area network. Not sure what you are experiencing, but I cannot confirm what you report. snip I cannot confirm that it is slower than VNC on a LAN but I can confirm that I have seen a dramatic increase in latency since the move to libssh. It still really feels like Nagle is enabled. I've not 100% ruled out that it is not something in my environment like packet loss on our connections but it doesn't feel like it. It feels exactly like packet coalescing as it is just the last bit of input that is badly delay - badly as in a second or so so very noticeable with serious production impact as it leads to dangerous mistakes while editing documents. I wonder if libssh does not have the ability to disable it. As a side note, I'm also really missing the setting of the ToS bits since moving to libssh. We currently have no way to distinguish X2Go interactive traffic (key strokes and screens) from bulk traffic (file transfers and print jobs). Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Full screen on Windows XP
On Mon, 2012-05-14 at 16:46 +0200, Paul van der Vlis wrote: Hello, Is it possible to work full-screen on Windows XP with X2goclient? I have a customer what says that this does not work. With regards, Paul. We have used it full screen. I do not recall the version I have installed on Windows right now and it is not immediately available. There are issues however. The keys to return from full screen do not work (e.g., CTL-M|F) nor does the magic pixel. The only way to exit full screen I found was to ALTTAB which reflects another problem. ALTTAB does not select windows in the virtual desktop as we would prefer it did. Is that what your client is experiencing? - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Temporary PulseAudio directory under `/tmp/` created by root
On Sat, 2012-04-28 at 10:56 +0200, Paul Menzel wrote: Dear X2go folks, I am trying to find out what package does try to connect to PulseAudio during system start up leaving a temporary directory `/tmp/pulse-` [1]. Could that be the result of running the init script `/etc/init.d/x2goserver`? This is a Debian Sid/unstable system. snip I haven't looked at the current code but I doubt it is /etc/init.d/x2goserver; that probably just invokes the session cleanup daemon. I do not recall which script it is but it is probably an x2go script in /usr/bin/ and invoked on the x2goserver via ssh from the client when sound is enabled - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] BUG REPORT - X2Go Client 3.99.2.1-rc2 does not work on Windows XP
On Wed, 2012-04-18 at 12:57 +0200, Stefan Baur wrote: Am 17.04.2012 12:28, schrieb Oleksandr Shneyder: snip I am unable to start the client on Windows XP (it works on Windows 7 and Vista): --- x2goclient.exe - Fehler in Anwendung --- Die Anwendung konnte nicht richtig initialisiert werden (0xc0150002). Klicken Sie auf OK, um die Anwendung zu beenden. --- OK --- snip Strange - I successfully installed and ran it on XP yesterday - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] [X2go-Dev] X2Go printing crash [was: libssh, nagle, and performance issues]
On Thu, 2011-09-22 at 07:51 +0200, Oleksandr Shneyder wrote: Am 21.09.2011 22:55, schrieb John A. Sullivan III: I tried to intercept it with ddd but that didn't work. Interestingly, its messages were that the program closed normally as if x2goprinting is invoking a close event. I'll see what I can do to get this to run from a debugger but it is a showstopping bug for us - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hallo John, I still thinking, that client crashes because it cannot work with output from your server. Can you give me a X2Go account on one of your servers, where you have troubles with printing? regards ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev It looks like the newer clients have magically solved the problem. Much appreciated! - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Traffic Shapping
On Thu, 2012-04-05 at 18:07 +0100, Helmer Teles wrote: Hello ML. Any of you guys have any experience in shaping x2go traffic to separate file printing sharing from normal session traffic? Any way to identify both print file tunnels from the rest of the ssh stream? snip Hi, Helmer. I'll paste in what I sent to you privately as I'd love to see us restore the old functionality if possible: We used to be able to distinguish interactive from bulk X2Go traffic when it was using OpenSSH. By default, SSH set the minimize-delay ToS bit (0x10)in the IP header for interactive traffic and minimize-cost (I think) (0x08) for bulk traffic. That seems to have gone away in the move to libssh instead of OpenSSH. I do hope they restore that differentiation so that we can shape the traffic differently. ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] [X2Go-User] Fwd: Flash Problems
On Sun, 2012-03-25 at 09:51 -0400, senrab...@aol.com wrote: Hi All: A quick follow up - we built out a 32 bit instance of Ubuntu Server 10.4 and re-tested - on a LAN, the video is pretty good - on a remote server (over the Internet), the results were like our 64bit experiencethe video stalls and refreshes so slowly it is pretty unusable. We thought the remote video worked well some time ago, though perhaps are wrong(?). Are there settings in x2go we should try, possibly on the compression method? Or any other thoughts? snip Our experience with NX for WAN video have been far less than satisfactory. We are eagerly waiting for SPICE to mature and hope it can be integrated as an alternative to NX into X2Go although I do see that SPICE development is leading toward implementing all the other features of X2Go as well such as sound, file sharing, and printing redirection - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] RDP services
I've been really looking forward to it but have not yet found the time. It will be important for clients where we have them booting from stick into a scaled down LXDE/XFCE environment. The big find for us and our clients using Windows XP desktops was the TSPlus Windows RDP7 client. I'm wondering what freerdp integration with X2Go would look like (which is, I assume, why you sent it to the dev list). Are you thinking that freerdp would be part of the x2goclient for Linux to connect to Windows RDP servers or that RDP would be an alternative protocol to NX for communication between X2Go client and server? If the latter, I've never looked seriously at the RDP server side solution for Linux but I thought it relied upon VNC which, I assume, would make it rather slow. If not, and it really does use fully native RDP6+, that could be an interesting solution. One problem we are seeing using RDP as a transport is no way to distinguish bulk from interactive RDP traffic for traffic shaping. That would be something we'd have to address. In fact, not to drift too far afield, but I think we have this problem in the current implementation of X2Go using SSH. I think the move to libssh has eliminated the practice of setting the minimize-latency TOS bit for interactive traffic and minimize-cost for bulk like openssh does. That has hurt us as we can no longer shape X2Go interactive traffic separately from bulk on our WAN/Internet connections. I have no idea how difficult that is to fix using libssh. Thanks - John On Sun, 2012-03-11 at 12:08 -0400, brian mullan wrote: John Have you tried the new FreeRDP client yet? I've not done any extensive testing but it does now support both Microsoft's RemoteFX RemoteApp services. Its an Apache license app and v1.0 was released and demo'd at the Ubuntu Developer conference in Nov 2011. RemoteFX - https://github.com/FreeRDP/FreeRDP/wiki/RemoteFX and RemoteApp services - https://github.com/FreeRDP/FreeRDP/wiki/RemoteApp I've configured Windows 2008 r2 servers on KVM and on AWS EC2, published applications for RemoteApp and then from linux/ubuntu machines run the xfreerdp client and the Windows RemoteApp appears on screen like any other app. I used to have to use seamlessrdp for this but now it just works since the freerdp v1.x release. I'd like to see FreeRDP support in x2go as I understand in the next Ubuntu FreeRDP is replacing rdesktop. Brian -- Forwarded message -- From: John A. Sullivan III jsulli...@opensourcedevel.com To: x2go-dev@lists.berlios.de Cc: Date: Sat, 10 Mar 2012 06:30:51 -0500 Subject: Re: [X2Go-Dev] X2Go Sessions File Format Specs On Sat, 2012-03-10 at 11:51 +0100, Mike Gabriel wrote: Hi Alex, hi all, I have started documenting the versatile session profile options of X2Go Client and Python X2Go (as found in ~/.x2goclient/sessions). @Alex: Could you please proof-read this wiki page and fix+change +amend? http://wiki.x2go.org/doku.php/wiki:development:sessions-file snip Thanks for documenting these, Mike; I read them with great interest. Toward the end, you write, Setting the command to RDP will turn X2Go into an RDP proxy (RDP via NX). RDP will receive a peformance improvement on WAN connections In our experience, this is true for RDP5 but it significantly degrades the native performance of RDP6 and RDP7. For those wishing to use RDP6 or 7 with Windows XP, TSPlus (www.tsplus.net) makes an XP RDP6/7 client and distributes it freely. I would love to see mimebox included in x2goclient. I know we had planned to write a multimedia player to automatically redirect to mimebox similar to the way printing works but have not gotten around to it - still overwhelmed by a large consulting contract completely outside of X2Go :( Thanks again - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] X2Go Sessions File Format Specs
On Sat, 2012-03-10 at 11:51 +0100, Mike Gabriel wrote: Hi Alex, hi all, I have started documenting the versatile session profile options of X2Go Client and Python X2Go (as found in ~/.x2goclient/sessions). @Alex: Could you please proof-read this wiki page and fix+change+amend? http://wiki.x2go.org/doku.php/wiki:development:sessions-file snip Thanks for documenting these, Mike; I read them with great interest. Toward the end, you write, Setting the command to RDP will turn X2Go into an RDP proxy (RDP via NX). RDP will receive a peformance improvement on WAN connections In our experience, this is true for RDP5 but it significantly degrades the native performance of RDP6 and RDP7. For those wishing to use RDP6 or 7 with Windows XP, TSPlus (www.tsplus.net) makes an XP RDP6/7 client and distributes it freely. I would love to see mimebox included in x2goclient. I know we had planned to write a multimedia player to automatically redirect to mimebox similar to the way printing works but have not gotten around to it - still overwhelmed by a large consulting contract completely outside of X2Go :( Thanks again - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] cups-x2go problem printing multiple copies from windows terminal server. Fix inside ( i guess )
On Thu, 2012-03-08 at 17:31 +0100, Mike Gabriel wrote: Hi Helmer, On Do 08 Mär 2012 16:26:31 CET Helmer Teles wrote: Steps to correct it: I've changed the line in /usr/lib/cups/backend/cups-x2go to: # PS2PDF command my $ps2pdf = /usr/bin/gs -q -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH - dSAFER -sDEVICE=pdfwrite -sOutputFile=\%s.pdf\ -dAutoRotatePages=/PageByPage -dAutoFilterColorImages=false -dColorImageFilter=/FlateEncode - dPDFSETTINGS=/printer -dDoNumCopies -c .setpdfwrite -f \%s\; FROM: # PS2PDF command my $ps2pdf = /usr/bin/gs -q -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH - dSAFER -sDEVICE=pdfwrite -sOutputFile=\%s.pdf\ -dAutoRotatePages=/PageByPage -dAutoFilterColorImages=false -dColorImageFilter=/FlateEncode - dPDFSETTINGS=/prepress -c .setpdfwrite -f \%s\; Patch applied, thanks! http://code.x2go.org/gitweb?p=cups-x2go.git;a=commitdiff;h=86e725ac8e514d40abcef9db2367be05736d0604 snip I do not know much about ghostscript but do we know what this change does and what the real problem was? Was it the PDFSETTINGS=/printer or the DoNumCopies? What does the change really do? A half hour of Internet research didn't turn up much: -dPDFSETTINGS=/screen (screen-view-only quality, 72 dpi images) -dPDFSETTINGS=/ebook(low quality, 150 dpi images) -dPDFSETTINGS=/printer (high quality, 300 dpi images) -dPDFSETTINGS=/prepress (high quality, color preserving, 300 dpi imgs) -dPDFSETTINGS=/default (almost identical to /screen) I do not know what color preserving means. I am a little concerned about size. We have had some issues where the wrong settings is the difference between sending a 2MB file and a 40MB file (I don't recall the exact number but I believe the difference was much greater) which can be critical when printing across a WAN. This is from the CutePDF FAQ: How to reduce the size of output PDF file? You can alter the parameters used in a text file %Program Files%\Acro Software\CutePDF Writer\PDFWrite.rsp. The key parameters are in the line: -dPDFSETTINGS=configuration where configuration can be /screen, /printer, /prepress, /default. Using /Printer can cut size by half. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] cups-x2go problem printing multiple copies from windows terminal server. Fix inside ( i guess )
On Thu, 2012-03-08 at 18:19 +0100, Mike Gabriel wrote: snip Thanks for your critical feedback. I have only applied the DoNumCopies setting, the PDFSETTINGS=prepress I have left untouched. I will build a package of the latest commit and Helmer has to test. Ok Helmer? Mike snip Thanks, Mike - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] cups-x2go problem printing multiple copies from windows terminal server. Fix inside ( i guess )
On Thu, 2012-03-08 at 21:56 +0100, newsgroups.ma...@stefanbaur.de wrote: Am 08.03.2012 20:25, schrieb John A. Sullivan III: I'm guessing the magic is removing DoNumCopies but that seems quite counter-intuitive. Uh, no, the patch is to add that option, not to remove it. Helmer's posting contained the patched version above the original. snip Argh! That should teach me to fly through emails while busy! Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Mac Client status
On Tue, 2012-03-06 at 18:21 +0100, Mihai Moldovan wrote: Hi Mike, * On 06.03.2012 09:48 AM, Mike Gabriel wrote: You should wait for 3.99.1.1, I guess. There is an X-Server detection fix (using unix socket, not TCP to connect to the local X-Server) in it. Probably. :) I do have a got git-snapshot with this fix in place (or could easily fetch the newest changes and rebuild), but this production-ready requirement makes me hesitate. ;) Mihai, is anything else needed to be patched into x2goclient.git currently for Mac OS X? Depends... here's a list of problems I FOUND (but not fixed): - sometimes, I have to resume a session, restart the client, suspend the session and resume it again for it to be usable (first time resuming will just close the connection to the server), BUT IIRC this happens on Linux, too, so I can't actually blame the OS X part... (plus, it only happens sometimes. :() - Client seems to be working fine for me on OS X 10.6/Snow Leopard. But: - has some weird memleak on OSX 10.5 (not always, again...) - Stephan reported problems on 10.7 which seem to be SSH-related, I've gotta take a look at that, no idea if I can figure it out though. Looks like a libssh problem, but at least I've got some debug output. No concrete code fixes from my side yet though. snip I have noticed that problem in Linux. I think it is an SSH timing issue. The connecting, suspending, reconnecting is not actually doing anything, I believe, except functioning as a rudimentary clock. If one just waits five minutes after being abruptly disconnected, all reconnects fine :( - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Mac Client status
On Tue, 2012-03-06 at 09:48 +0100, Mike Gabriel wrote: Hi John, On Di 06 Mär 2012 01:06:55 CET John A. Sullivan III wrote: Hello, all. We were delighted to see Mihai's find about the overflow as we do have client whose Mac always freezes if the screen saver kicks in. We are eager to upgrade her but are unsure of the current state of the Mac client. Is there a production ready version with this fix available? If so, which one? Thanks - John You should wait for 3.99.1.1, I guess. There is an X-Server detection fix (using unix socket, not TCP to connect to the local X-Server) in it. Mihai, is anything else needed to be patched into x2goclient.git currently for Mac OS X? Thanks, Mike snip Thanks very much, Mike. We'll eagerly await the new release - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Mac Client status
On Tue, 2012-03-06 at 22:55 +0100, Mike Gabriel wrote: Hi guys, On Di 06 Mär 2012 18:48:12 CET John A. Sullivan III wrote: On Tue, 2012-03-06 at 18:21 +0100, Mihai Moldovan wrote: - sometimes, I have to resume a session, restart the client, suspend the session and resume it again for it to be usable (first time resuming will just close the connection to the server), BUT IIRC this happens on Linux, too, so I can't actually blame the OS X part... (plus, it only happens sometimes. :() snip I have noticed that problem in Linux. I think it is an SSH timing issue. The connecting, suspending, reconnecting is not actually doing anything, I believe, except functioning as a rudimentary clock. If one just waits five minutes after being abruptly disconnected, all reconnects fine :( - John When writing the connection/tunneling code of python-x2go I observed something similar. The problem with port forwarding requests (both reverse or non-reverse) is a situation when (after suspend) the ports do no get closed properly. We always have to make sure to properly cancel port forwarding requests before we disconnect from a SSH session. snip So how should we respond if the network connection is broken uncleanly, e.g., the WAN or Internet connection drops or the client computer crashes? ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Mac Client status
On Tue, 2012-03-06 at 23:12 +0100, Mike Gabriel wrote: Hi John, On Di 06 Mär 2012 23:00:35 CET John A. Sullivan III wrote: On Tue, 2012-03-06 at 22:55 +0100, Mike Gabriel wrote: Hi guys, On Di 06 Mär 2012 18:48:12 CET John A. Sullivan III wrote: On Tue, 2012-03-06 at 18:21 +0100, Mihai Moldovan wrote: - sometimes, I have to resume a session, restart the client, suspend the session and resume it again for it to be usable (first time resuming will just close the connection to the server), BUT IIRC this happens on Linux, too, so I can't actually blame the OS X part... (plus, it only happens sometimes. :() snip I have noticed that problem in Linux. I think it is an SSH timing issue. The connecting, suspending, reconnecting is not actually doing anything, I believe, except functioning as a rudimentary clock. If one just waits five minutes after being abruptly disconnected, all reconnects fine :( - John When writing the connection/tunneling code of python-x2go I observed something similar. The problem with port forwarding requests (both reverse or non-reverse) is a situation when (after suspend) the ports do no get closed properly. We always have to make sure to properly cancel port forwarding requests before we disconnect from a SSH session. snip So how should we respond if the network connection is broken uncleanly, e.g., the WAN or Internet connection drops or the client computer crashes? What I do is, I try to connect and if that fails, I try to cancel the port forwarding requests. This mostly works and sometimes also fails. snip Yes, but that is what we do as users. What can we do as developers to detect and correct this situation so users do not have to do this? Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Preferred settings for X2go LAN use
On Wed, 2012-02-22 at 09:30 +0100, Moritz Strübe wrote: Am 21.02.2012 20:36, Michael Kromer schrieb: I would recommend you to use jpeg compression in combination with libjpegturbo A while ago I did some experiments, too. I ended up using RGB. Even using RLE put significant more load on the system I was testing (Server and Client). Due to the artefacts JPEG was never an option to me. But probably you won't get around some experiments, because there are just too many things that influence the results (bandwidth, delay, server architecture, client architecture, etc). We should start a wiki-page on this. ;) snip Interesting. I have not had the time to pursue it but, I found in one particular installation where we were running X2Go from a Windows RDS session, anything that used image compression was completely corrupted - jpeg and png. We had to revert to RGB. I do not know why - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Traffic shaping nightmare - missing SSH TOS bits
On Tue, 2012-02-21 at 09:30 +0100, Jan Engelhardt wrote: On Tuesday 2012-02-21 07:05, John A. Sullivan III wrote: I've been working on a project to implement traffic shaping to conform to 95th percentile billing including our X2Go virtual desktops. One of the challenges is prioritizing the interactive X2Go traffic (screen, keyboard, mouse) differently from bulk traffic like local share transfers, printing, and possibly sound. I thought this would be trivial because of the SSH transport since SSH sets the minimize-delay bit for interactive traffic and the maximize-throughput bit for bulk transfers. If a user has his own machine, he can of course do whatever he wants. Second, if he can pass options to ssh, he can set the QOS type arbitrarily, and even where that is not available for control, one can still cause the openssh client to use interactive TOS marking for bulk transfers, since secondary ssh channels (e.g. tunnels) do not pass up the TOS mark. snip However, Windows and Mac clients may not be using OpenSSH. Users may not have control over their X2Go Server configurations; they certainly do not in our environment. We do need a way to distinguish between bulk and interactive traffic to ensure adequate performance in any bandwidth constrained environment. Thanks - John PS - out of curiosity, now does one set the QoS type in OpenSSH? Several minutes of searching has not given me an answer. That's far from an exhaustive search but, if someone knows off hand, it would be helpful. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Traffic shaping nightmare - missing SSH TOS bits
On Tue, 2012-02-21 at 10:11 +0100, Jan Engelhardt wrote: On Tuesday 2012-02-21 09:43, John A. Sullivan III wrote: On Tue, 2012-02-21 at 09:30 +0100, Jan Engelhardt wrote: On Tuesday 2012-02-21 07:05, John A. Sullivan III wrote: I've been working on a project to implement traffic shaping to conform to 95th percentile billing including our X2Go virtual desktops. One of the challenges is prioritizing the interactive X2Go traffic (screen, keyboard, mouse) differently from bulk traffic like local share transfers, printing, and possibly sound. I thought this would be trivial because of the SSH transport since SSH sets the minimize-delay bit for interactive traffic and the maximize-throughput bit for bulk transfers. If a user has his own machine, he can of course do whatever he wants. Second, if he can pass options to ssh, he can set the QOS type arbitrarily, and even where that is not available for control, one can still cause the openssh client to use interactive TOS marking for bulk transfers, since secondary ssh channels (e.g. tunnels) do not pass up the TOS mark. snip However, Windows and Mac clients may not be using OpenSSH. That too. All these reasons is why TOS is not a reliable characteristic to determine the question of bulk-or-interactive. PS - out of curiosity, now does one set the QoS type in OpenSSH? ssh -o IPQoS={bulk|interactive|0x12345} ... ssh -L 2000:localhost:22 foohost rsync -HPave 'ssh -p 2000' bulky.iso localhost: and in fact, wouldn't the output `ls -Rl /` in an interactive shell also best be count as bulk... snip Interesting. Looks like that was introduced in 5.6. Squeeze is running 5.5 but it should default to 0x10/0x08 if not set. I'm not seeing any TOS bits set which is why I suspect it is the libssh call as I assume that does not reference the openssh configuration. If someone can point out where the socket options are set, I can take a look and see if I can figure out how to fix it but I don't know a thing about socket programming. I assume one has to make some kind of a call to IP_TOS or perhaps it is part of setsockopts and SO_PRIORITY. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Traffic shaping nightmare - missing SSH TOS bits
On Tue, 2012-02-21 at 18:22 +0100, Mike Gabriel wrote: Hi John, On Di 21 Feb 2012 17:09:57 CET John A. Sullivan III wrote: On Tue, 2012-02-21 at 10:11 +0100, Jan Engelhardt wrote: On Tuesday 2012-02-21 09:43, John A. Sullivan III wrote: On Tue, 2012-02-21 at 09:30 +0100, Jan Engelhardt wrote: On Tuesday 2012-02-21 07:05, John A. Sullivan III wrote: I've been working on a project to implement traffic shaping to conform to 95th percentile billing including our X2Go virtual desktops. One of the challenges is prioritizing the interactive X2Go traffic (screen, keyboard, mouse) differently from bulk traffic like local share transfers, printing, and possibly sound. I thought this would be trivial because of the SSH transport since SSH sets the minimize-delay bit for interactive traffic and the maximize-throughput bit for bulk transfers. If a user has his own machine, he can of course do whatever he wants. Second, if he can pass options to ssh, he can set the QOS type arbitrarily, and even where that is not available for control, one can still cause the openssh client to use interactive TOS marking for bulk transfers, since secondary ssh channels (e.g. tunnels) do not pass up the TOS mark. snip However, Windows and Mac clients may not be using OpenSSH. That too. All these reasons is why TOS is not a reliable characteristic to determine the question of bulk-or-interactive. PS - out of curiosity, now does one set the QoS type in OpenSSH? ssh -o IPQoS={bulk|interactive|0x12345} ... ssh -L 2000:localhost:22 foohost rsync -HPave 'ssh -p 2000' bulky.iso localhost: and in fact, wouldn't the output `ls -Rl /` in an interactive shell also best be count as bulk... snip Interesting. Looks like that was introduced in 5.6. Squeeze is running 5.5 but it should default to 0x10/0x08 if not set. I'm not seeing any TOS bits set which is why I suspect it is the libssh call as I assume that does not reference the openssh configuration. If someone can point out where the socket options are set, I can take a look and see if I can figure out how to fix it but I don't know a thing about socket programming. I assume one has to make some kind of a call to IP_TOS or perhaps it is part of setsockopts and SO_PRIORITY. Thanks - John The sockets are probably created and configured in here... http://code.x2go.org/gitweb?p=x2goclient.git;a=blob;f=sshprocess.cpp Mike snip Thanks. What if it is the unusual case where they are not tunneling through ssh? - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] x2goclient.exe - Saving session settings to registry
On Tue, 2012-02-21 at 22:19 +0100, newsgroups.ma...@stefanbaur.de wrote: Am 21.02.2012 22:09, schrieb Heinz-M. Graesing: Hello Stefan, Am 21.02.2012 15:51, schrieb newsgroups.ma...@stefanbaur.de: Hi List, Using the sessions file is the default for Linux and AFAIK also for Mac OS X, so why the variation in x2goclient.exe? The registry db is the official recommended place to store application settings on this platform. This means that people (and applications) working with this system are looking for such information inside the registry - not an ini file. There is no way to preseed a user profile, though, aside from some hacks using reg.exe and a login script, as everything gets written to HKEY_CURRENT_USER. You cannot set a domain- or machine-specific default/template (the User named .Default does not what one might think it does ;-)) . Also, once it is written to HKEY_CURRENT_USER, the user can change it at will, there is no way to set it to read-only (while there is --no-session-edit, there's nothing to stop the user from clicking on x2goclient.exe directly). Then again, the problem should become moot in about three weeks time, when selectively publishing applications becomes available. ;-) As this option is already is introduced I think a second option can only be introduced as optional. Or, you could add an autosensing capability: If there's a sessions file present, pick that, if not, write to the registry. That way, copying a sessions file / touching an empty file in the default location would tell x2goclient that it should refrain from storing data in the registry. snip I'm a Windows ignoramus but is this something that could be set with a GPO? - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
[X2Go-Dev] Traffic shaping nightmare - missing SSH TOS bits
Hello, all. I've been working on a project to implement traffic shaping to conform to 95th percentile billing including our X2Go virtual desktops. One of the challenges is prioritizing the interactive X2Go traffic (screen, keyboard, mouse) differently from bulk traffic like local share transfers, printing, and possibly sound. I thought this would be trivial because of the SSH transport since SSH sets the minimize-delay bit for interactive traffic and the maximize-throughput bit for bulk transfers. Thus we could distinguish the packets by looking at the TOS bits. This is how the default Linux pfifo_fast qdisc does it anyway. I took a packet trace just to confirm it and, to my horror, I see that we are not setting any of the TOS bits and thus not prioritizing interactive traffic even with the default traffic shaping. I'm guessing this is a casualty of moving to libssh. Is there any way to set the TOS bits for interactive to 0x10, bulk to 0x08 as it is in OpenSSH?. It would be really nice if sound could be configured either way according the the administrator's preference. Unfortunately, I did not see anything in the libssh documentation to show how to do this although I do not claim to understand it. This is a critical issue for any WAN based deployments like ours. Thanks - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Font colour in x2goclient in case of dark system theme
On Fri, 2012-02-17 at 16:33 +0100, Milan Knížek wrote: Hello, the x2goclient has hardcoded light appearance for the various widgets (main window, session list, login window, progress of logging in), which does not work well with dark system theme: the white font is not distinquishable from the background. I have checked the Qt documentation and there seems to be more ways on how to accomplish that. So far, I have used three various methods: labelname-setStyleSheet(QLabel { color : black; }); or cpal.setColor ( QPalette::Text, Qt::black); or QString fonttemplate=(font color='black'%1/font); nameLabel-setTextFormat ( Qt::RichText ); nameLabel-setText ( fonttemplate.arg( text ) ); which kind of works, though it requires to be written repeatedly throughout the code (sessionbutton.cpp, onmainwindow.cpp, onmainwindow_part4.cpp) for individual widgets and so far I am not able to use a single approach for all widget types (QLabel, QMenu, QPushButton, etc.). The first method seems to be the most general and simplest, however. Is there a way to set application-wide font colour for anything that appears in the widgets, but not in app menu and other windows (Settings, Session Settings, etc.), because these are styled with the system-wide dark theme? Where in the code? P.S. Sorry for possibly stupid question, but since I am not a coder, it is rather a trial error approach and a lot of googling... snip I don't think it is stupid at all. I have found myself sometimes frustrated at white text boxes with no borders on white backgrounds making it almost impossible to see where to click to enter the field. Thanks for bringing it to the fore - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] ATTN: Alex - Questions regarding x2goclient.exe --portable and session specification
On Thu, 2012-02-16 at 09:38 +0100, Oleksandr Shneyder wrote: Am 16.02.2012 09:31, schrieb newsgroups.ma...@stefanbaur.de: Am 16.02.2012 09:28, schrieb Oleksandr Shneyder: I was hoping I could dump the sessions file into a central folder, say, c:\program files\x2goclient\, but that won't work if cache files will be stored in the same directory. I don't think that it's good idea. You can create a file with session settings and copy it to home folder of each user That becomes a royal pain in the lower back when you have to change something in the session file... Sorry, in my opinion the session settings are very user specific and should be not saved centrally. You can write a simple shell script to copy template file with session settings into directories of each user on system. snip I wonder if a hierarchical approach is in order as many of the settings may be completely the same for all users and it would be desirable to centrally manage them. The KDE KIOSK is a good model where settings can be set centrally, environment variables can be substituted, settings can be locked, but, if they are not locked, they can be overridden locally with the most local version taking priority. Just a thought. I share Stefan's pain :) - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] [X2go-User] Problem with xfreerdp ( any version )
On Tue, 2012-02-14 at 12:06 +0100, Mike Gabriel wrote: Hi Alex, Helmer, I have forwarded the mail for freerdp-devel, let's see if I am lucky. I am subscribed to the list so there is a chance... Mike snip I hope so. This would definitely affect us so thanks very much for taking the time to fix it even though it was not an X2Go problem - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] Review: changes and features in X2Go
On Thu, 2012-02-02 at 11:56 +0100, Oleksandr Shneyder wrote: Hello X2Go users, I want to describe some changes we made in X2Go recently. == Multi display support features. == There are two new features in X2Go to support desktops with more then one physical display. 1. Run X2Go Session in full screen mode on one of several physical displays. With this feature you can use one of your physical monitors to display X2Go session in full screen mode. On this photo you can see a Windows 7 notebook with a connected VGA monitor running X2Go Session on display 2: http://www.x2go.org/_media/news:two_sess.png You can configure this feature using Session Preferences dialog in X2Go client: http://www.x2go.org/_media/news:multidisp_set.png Push Identify all displays button to find out the numbers of a physical displays: http://www.x2go.org/_media/news:ident_dsip.png 2. Using two or more physical displays as one large virtual display (Xinerama) The advantage of Xinerama feature is to only maximize windows to the dimensions of the active physical display. This can be very useful not only for thinclients with more then one physical displays, but also if user stretch their X2Go sessions over more then one display in window mode. X2Go Session running 3 maximized LibreOffice windows is stretched over 3 physical displays: http://www.x2go.org/_media/news:xinerama.png Xinerama configuration of X2Go session will be automatically updated every time when user moving or resizing his X2Go Session window. ==Verifying if port is free before resuming X2Go Session == On start of an X2Go session 3 tcp ports for ssh tunnels (graphic, sound, sshfs) are reserved. The values of these ports are stored in the X2Go session database and used by X2Go clients every time an X2Go session gets resumed. Sometimes it can happen that one of these ports is already used by another application after the session had been suspended. As result, X2Go client will be unable to resume such session or start sound or sshfs tunnel. To avoid this problem X2Go server (= 3.1.0.0) will verify if these ports are still unused before the session will be resumed. In other case new ports will be reserved. The new port values will be updated in database and sent to the X2Go client. IMPORTANT: Older versions of X2Go client will be compatible with newest X2Go servers. If ports have changed between suspend and resume of a session, older x2goclients will try to reconnect and fail. But on the next attempt, X2Go client will retrieve the updated port values from the X2Go session database. snip Very much appreciated. Thank you - John ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2Go-Dev] x2goclient entered Debian unstable
On Sat, 2012-01-28 at 10:02 +0100, Angela Fuß wrote: WOW Congratulations! I am impressed. snip +1 ___ X2Go-Dev mailing list X2Go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] replace for kde/lxde/gnome-bindings and x2godesktopsharing
On Fri, 2012-01-27 at 09:16 +0100, Oleksandr Shneyder wrote: Hello dear X2Go users, I want to share with you my idea to create new application, which should replace such packages as kde/lxde/gnome-bindings and x2godesktopsharing. The reasons to create such application are: - reduce effort of developing and supporting bindings for every desktop environment - put all x2go related settings in one place. That should improve usability - add some new features The application should provide existing features: - display and manage(open, demount) exported shares - manage sharing of display (x2godesktopsharing) if display sharing is configured - suspend X2Go session The application should be able to communicate with x2goclient and provide some new features: - switch between full screen and window modes - enable/disable multi display (Xinerama) support - send x2goclient command to open export directory dialog. This should be very useful if X2Go session is running in full screen mode or x2goclient is minimized to system tray. I'll be appreciative for your comments and for suggestions about this application's name. snip I think it is an excellent idea. Of course, the devil will be in the details :) - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] RPMs for RH6
On Fri, 2012-01-27 at 15:55 +0100, Mike Gabriel wrote: Hi Bogdan, snip Q2: Why do you configure Postgres? For packaging please provide an SQLite solution. X2Go can be used without Postgres and without Postgres it is possible to install X2Go server and let it run out of the box. Please pre-configure X2Go server with SQLite. Please refer to the postinst script in the /debian folder: http://code.x2go.org/gitweb?p=x2goserver.git;a=blob;f=debian/x2goserver.postinst;h=68d83d755e683fc9e04e9b5d36ca046c01e13bb0;hb=ce3023c2ee41457a4acf19ec858c84ddf068ee1f snip Shouldn't that be an option? I would think PostgreSQL would be the preferred approach for larger or multiserver installations - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Informations about sshfs and mounting local device
On Thu, 2011-12-22 at 12:10 +0100, Mike Gabriel wrote: Hi Jean, On Do 22 Dez 2011 11:37:15 CET Jean Milot wrote: Hi Mike, Yes, i use Freenx Server for now. And, i would like to know how it works for X2Go. What have you done and what problems have you had ? Thanks for your help. The handshake for SSHFS mounting (server - client) is: o client sets up a reverse SSH port forwarding tunnel so that you can login from server to client, even if the client is behind a firewall o client generates temporary SSH pub/priv key pair o a script on the server is then issued by the client o with this script the client places this temp SSH privkey and a known_hosts file that works for the client-side sshd on the server o from this script also sshfs mounting (server - client) is issued o once the mount has been successfull, all temporary credentials are removed from the server again (as I said: temporary) You might want to take a closer look at the ,,x2gomountdirs'' script in X2Go server and the appropriate locations in x2goclient or python-x2go. Tell me what language (C++ or Python) you prefer for code diving and I will point you to the right places in the code. Snip If I recall correctly, there is also an issue with some distributions (particularly Ubuntu) where doing the mount provokes a new media dialog for all NX users connected to the server. I believe this is why X2Go actually does the mount in /tmp and then makes a symbolic link back to the user's home directory - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] SOLUTION: A generic way to print to cheap GDI-only printers (not only with X2Go)
On Wed, 2011-12-21 at 00:37 +0100, Stefan Baur wrote: Hi List, since I've had issues getting client-side printing to work (which I need for low-budget printers that don't support PCL/PS), I built my own work-around. snip PS: I'd still be interested in knowing what's currently going wrong with the X2Go-client's built-in printing support on Windows (see my post at https://lists.berlios.de/pipermail/x2go-dev/2011-December/003133.html). Should be way less of a hassle to configure once it works. snip Hi, Stefan. At some point, I'll have to have a close look at this myself. Windows printing is working for us but we are using a very old client. We have not invested the time to work through the new client but there do appear to be several issues with it. I hope to be able to dive in within a couple of months - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Printing-related question for Windows X2go client users
On Sat, 2011-12-10 at 14:36 +0100, Stefan Baur (newsgroups-Kontakt) wrote: Hi list, this question goes out to all the Windows users of the X2go-Client: Has anybody ever tried to use Sumatra PDF as a PDF endpoint on the Windows client, instead of Adobe Reader? It seems rather interesting, as it is GNU licensed, offers commandline options to print a file to the default printer (or show a print dialog) and auto-terminate itself after printing has finished (look for Command-line arguments in the online manual linked below). Info in English: http://en.wikipedia.org/wiki/Sumatra_PDF http://blog.kowalczyk.info/software/sumatrapdf/manual.html Info in German: http://de.wikipedia.org/wiki/Sumatra_PDF http://blog.kowalczyk.info/software/sumatrapdf/manual-de.html I'd love to try it out myself, but I haven't managed to get printing/showing a PDF on the Client to work at all so far, thus testing Sumatra PDF doesn't really make sense on my client yet. So here's to crowdsourcing - anybody ever tried it, anybody with a working client printer setup willing to try? snip It looks very interesting and we'd be interested in trying it. However, PDFs have been a real sore spot in our VDI work and our offering both Linux and Windows desktops for those who would like to leave the Windows world for any of several reasons. Not only is there nothing even close to Acrobat for editing PDFs in the open source world, the readers also have serious shortcomings. Believe me, I am no Adobe fan and wish things were different. One of our clients is a construction company who handle very large numbers of very large PDFs - their construction drawings. Last we tested, there were serious bugs in every single non-Adobe PDF reader on Linux which refused to properly print the larger drawings. They always shrunk them to the default paper size. Moreover and more to the point at hand, none of the open source PDF readers were able to support all the features we frequently encountered among these and other documents. If the reader only handles 85% of documents and has trouble printing, one is required to keep the Acrobat Reader around. Most of our users would prefer to not deal with two applications. I suspect that Sumatra, being lightweight by design, will fall into the same category. Not that Acrobat Reader is not without problems. Besides its bloat, the Linux reader has hard coded the permissions of set files to 644 wreaking havoc on our shared directories where everything is setgid with a 770 umask so that shared documents can be edited by the group. So, the point of it all is that I suspect there will be problems using Sumatra - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Printing-related question for Windows X2go client users
On Sat, 2011-12-10 at 17:28 +0100, Stefan Baur (newsgroups-Kontakt) wrote: John A. Sullivan III wrote: On Sat, 2011-12-10 at 14:36 +0100, Stefan Baur (newsgroups-Kontakt) wrote: Hi list, this question goes out to all the Windows users of the X2go-Client: Has anybody ever tried to use Sumatra PDF as a PDF endpoint on the Windows client, instead of Adobe Reader? snip It looks very interesting and we'd be interested in trying it. However, PDFs have been a real sore spot in our VDI work and our offering both Linux and Windows desktops for those who would like to leave the Windows world for any of several reasons. Not only is there nothing even close to Acrobat for editing PDFs in the open source world, the readers also have serious shortcomings. further PDF/Adobe sucks but we'll just have to deal with it rant, *that I can fully relate to*, snipped I can see why one would not want to use a non-Adobe PDF viewer as regular PDF viewer, given the numerous extensions to the original PDF spec that may or may not be supported by such a third-party client, but my intent here was to (ab)use Sumatra PDF for printing via the Windows X2go client *only*. I am *not* suggesting replacing the default PDF viewer on the client computers, nor on the host (which wouldn't work with Sumatra PDF anyway, as it is a Windows-only application). The reasoning why I'd prefer printing via a PDF transferred to the client is the following: - Linux and Mac OS are Postscript-centered when it comes to printing, and whoever is running a Linux client usually also knows what printer to buy (i.e. one with Postscript or PCL support). So there is no need for a third-party solution here. - In the world of Microsoft Windows, however, such nasties as GDI-only printers exist. These are usually low-priced printers that cannot be made to work with Linux or Mac OS. Converting a Postscript printjob into a PDF is easy, though (that's what CUPS does when you install the virtual PDF printer) - and passing the resulting PDF file to a viewer like Sumatra PDF that automatically starts to dump it into the proper printer queue on the client means: We have a printing solution that does *not* require some sort of Linux driver for the printer, letting us use all the fancy features the printer may have that are supported by its Windows driver (think duplexing or different paper trays/outbins - these are usually a pain in the behind to get to work when all you have is a generic PCL-something PPD, or color profiles). My general assumption behind this is that if you try to print a PDF from Adobe Reader for Linux on the host, it will be converted into Postscript first, then into whatever printer-specific language that may be required: usually PCL 3, PCL5, PCL 6 - or back into PDF. This conversion from PDF to PS and again into PDF with CUPS is, from my experience, *not* a 1:1 mapping/translation, so the resulting PDF should be a basic PDF type with all the trouble-causing fanciness removed (and usually resulting in a significantly larger spool file). Of course, it's been at least three years, if not more, since I last had to deal with conversion of printing formats, so, If I'm wrong here and my knowledge is outdated, or your experience proves me wrong, please feel free to correct me. Also, to quote from Sumatra PDF's Wikipedia entry: It has a 4.4 MB setup file, compared to Adobe Reader's 40.5 MB, for Windows 7. Installed size is 8.4 MB, whereas Adobe Reader requires 335 MB of available disk space. ==so I wouldn't see disk space/size/bloat issues. Plus, since it is open source, it might even be possible to integrate the relevant parts of its code into the X2go client itself - any of the coders willing to comment on the feasibility of that? If that works, the end user wouldn't even get to see a second PDF viewer program, as it would become part of the X2go client's code. snip I admit I just scanned through your response (deadlines :( ) but I think that is the way X2Go printing is working. The x2go cups driver is converting to pdf and sending the pdf to the client. In the Linux and Mac clients, we can print directly to the printer but, in the Windows client, it always opens the default PDF reader no matter which printer one chooses. That's why I assumed you were talking about replacing Acrobat as the default PDF reader, i.e., because as far as I know, that's what the x2go client prints to - the default PDF reader - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] [PATCH] Make sure the [...] - The right patch!
On Fri, 2011-12-09 at 18:58 +0100, Moritz Struebe wrote: Sorry for that, too many patches flying around. :) Morty On 2011-12-09 18:57, Moritz Struebe wrote: Hi there, this did cost us some pains, because a sshd was using port 6050, so even reinstalling x2goserver didn't solve any issues. This is a pretty hackish solution and maybe someone with a little more bash experience could take a look. snip Only two silly thoughts. Are we sure that all distributions use a space and not a \t after the port number? If not, perhaps something like netstat -ntl | grep :${X2GO_PORT}[^0-9][^0-9]* would work (the port followed by at least one non-numeric character. What about the bizarre corner case where someone may have specified a source port for the connection and that port happens to be ${X2GO_PORT}. Does that mean it will show up in the Foreign Address column? Would it be better to use awk to pull out the fourth column (I forget off hand how to do that) and then grep only the fourth column? Then we might even be able to use netstat -ntl | grep :${X2GO_PORT}\$ (not 100% sure of that syntax either). Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] [PATCH] Make sure the [...] - The right patch!
On Fri, 2011-12-09 at 21:53 +0100, Reinhard Tartler wrote: On Fr, Dez 09, 2011 at 18:58:35 (CET), Moritz Struebe wrote: --- a/x2goserver/bin/x2gostartagent +++ b/x2goserver/bin/x2gostartagent @@ -120,8 +120,8 @@ USED_DISPLAYS=`$X2GO_LIB_PATH/x2gogetdisplays $HOSTNAME` while [ $OUTPUT != inserted ]; do X2GO_PORT=$(($X2GO_PORT + 1)) X2GO_PORT=`echo for(\\$i=$X2GO_PORT;\\$br ne \true\;\\$i++){ if(\$USED_DISPLAYS\ =~ m/\\|\\$i\\|/){\\$br=\false\;}else{\\$br=\true\;print \\$i;}}|perl` - - if [ -e /tmp/.X${X2GO_PORT}-lock ] || [ -e /tmp/.X11-unix/X${X2GO_PORT} ]; then + #Test if the session is already in use. nxagent uses 6000+DISPLAY to open a port. Therefore this must be tested, too. + if [ -e /tmp/.X${X2GO_PORT}-lock ] || [ -e /tmp/.X11-unix/X${X2GO_PORT} ] || `netstat -ntl | grep :${X2GO_PORT} ` ; then OUTPUT=XXX else SESSION_NAME=${USER}-${X2GO_PORT}-`date +%s` That looks overly convoluted to me. What about this (untested, but shell syntax works): if [ -e /tmp/.X${X2GO_PORT}-lock ] || [ -e /tmp/.X11-unix/X${X2GO_PORT} ] || netstat -ntl 2/dev/null | grep -q :${X2GO_PORT}; then OUTPUT=XXX else SESSION_NAME=${USER}-${X2GO_PORT}-`date +%s` fi Cheers, Reinhard What if X2GO_PORT is 6051 and for some reason there is a service listening on port 60513? Won't it erroneously match? - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Problem with kde trinity
- Original Message - From: Mike Gabriel mike.gabr...@das-netzwerkteam.de To: x2go-dev@lists.berlios.de Sent: Thursday, November 3, 2011 4:31:52 AM Subject: Re: [X2go-Dev] Problem with kde trinity Hi Jean, On Do 03 Nov 2011 09:27:05 CET Jean Milot wrote: Hi, I would like use x2go with kde trinity. But it seems that x2gotrinitybindings call an old package. Or maybe i yave a problem with my source list. Thanks for your help. Jean Trinity support is only implement in the pyhoca-gui client, not in the native x2goclient. The Trinity implementation into x2goclient should be not problem at all. But: the x2gotrinitybindings package is not at all usable... It simply has been adopted from the old KDE3 x2gokdebindings package. x2gotrinitybindings (same as its predecessor x2gokdebindings) actually contains a bad hack that is not Debian policy conform: it moves a file installed by another package out of the way. Maybe this is where your development efforts could start? I know that John Sullivan (also on this list) deploys Trinity in cloud desktop computing, as well. So maybe you can bundle efforts. snip Yes, that is what we did - we renamed the /usr/bin/startkde and replaced it with a link to /opt/trinity/bin/startkde - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Problem with kde trinity
Certainly. Feel free to email me directly - John - Original Message - From: Jean Milot milot.j...@gmail.com To: x2go-dev@lists.berlios.de Sent: Thursday, November 3, 2011 3:49:03 PM Subject: Re: [X2go-Dev] Problem with kde trinity Hello John, I am interested in what you do. Maybe we can find a way to communicate more efficiently in order to share some informations and work together. 2011/11/3 John A. Sullivan III jsulli...@opensourcedevel.com - Original Message - From: Mike Gabriel mike.gabr...@das-netzwerkteam.de To: x2go-dev@lists.berlios.de Sent: Thursday, November 3, 2011 4:31:52 AM Subject: Re: [X2go-Dev] Problem with kde trinity Hi Jean, On Do 03 Nov 2011 09:27:05 CET Jean Milot wrote: Hi, I would like use x2go with kde trinity. But it seems that x2gotrinitybindings call an old package. Or maybe i yave a problem with my source list. Thanks for your help. Jean Trinity support is only implement in the pyhoca-gui client, not in the native x2goclient. The Trinity implementation into x2goclient should be not problem at all. But: the x2gotrinitybindings package is not at all usable... It simply has been adopted from the old KDE3 x2gokdebindings package. x2gotrinitybindings (same as its predecessor x2gokdebindings) actually contains a bad hack that is not Debian policy conform: it moves a file installed by another package out of the way. Maybe this is where your development efforts could start? I know that John Sullivan (also on this list) deploys Trinity in cloud desktop computing, as well. So maybe you can bundle efforts. snip Yes, that is what we did - we renamed the /usr/bin/startkde and replaced it with a link to /opt/trinity/bin/startkde - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev -- MILOT Jean Tél. : 0659514624 milot.j...@gmail.com ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Problem with windows x2go client 3.99.0.1
On Sat, 2011-10-15 at 20:38 +0200, va...@spsmb.cz wrote: Hello, x2go client 3.99.0.1 (and 3.99.0.0 and 3.01.18) is not working under windows 7. Session is established but connection timeout. On the server side in log: sshd[1903]: error: connect_to localhost port 30001: failed. I use debian squeeze and x2go server from http://packages.x2go.org/debian. 3.01.18 under ubuntu with no problem Where is the problem? snip I've noticed the exact same thing. I assumed it was because I am using 3.0.1-5 on the server side and there was a backwards compatibility issue so I postponed investigation until I have a chance to upgrade the server side - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Problem with windows x2go client 3.99.0.1
On Sat, 2011-10-15 at 22:40 +0200, va...@spsmb.cz wrote: Dne 15.10.2011 20:56, John A. Sullivan III napsal(a): On Sat, 2011-10-15 at 20:38 +0200, va...@spsmb.cz wrote: Hello, x2go client 3.99.0.1 (and 3.99.0.0 and 3.01.18) is not working under windows 7. Session is established but connection timeout. On the server side in log: sshd[1903]: error: connect_to localhost port 30001: failed. I use debian squeeze and x2go server from http://packages.x2go.org/debian. 3.01.18 under ubuntu with no problem Where is the problem? snip I've noticed the exact same thing. I assumed it was because I am using 3.0.1-5 on the server side and there was a backwards compatibility issue so I postponed investigation until I have a chance to upgrade the server side - John snip I use most recent x2go server (3.0.99.6). Only 3.01-4 client from http://www.x2go.org/index.php?id=7 works (3.01-2 should work but there are trojans inside of installation package). snip Hmm . . . smells like libssh. I wonder if it's related to another problem we are seeing in Linux with the newer clients; we are constantly dropping SSH sessions. Sound, file sharing, printing, all those functions which rely upon the reverse ssh forwarding all break after a while. Suspending and resuming the session restores them - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Problem with windows x2go client 3.99.0.1
On Sun, 2011-10-16 at 03:52 +0200, va...@spsmb.cz wrote: Dne 15.10.2011 23:01, John A. Sullivan III napsal(a): On Sat, 2011-10-15 at 22:40 +0200, va...@spsmb.cz wrote: Dne 15.10.2011 20:56, John A. Sullivan III napsal(a): On Sat, 2011-10-15 at 20:38 +0200, va...@spsmb.cz wrote: Hello, x2go client 3.99.0.1 (and 3.99.0.0 and 3.01.18) is not working under windows 7. Session is established but connection timeout. On the server side in log: sshd[1903]: error: connect_to localhost port 30001: failed. I use debian squeeze and x2go server from http://packages.x2go.org/debian. 3.01.18 under ubuntu with no problem Where is the problem? snip I've noticed the exact same thing. I assumed it was because I am using 3.0.1-5 on the server side and there was a backwards compatibility issue so I postponed investigation until I have a chance to upgrade the server side - John snip I use most recent x2go server (3.0.99.6). Only 3.01-4 client from http://www.x2go.org/index.php?id=7 works (3.01-2 should work but there are trojans inside of installation package). snip Hmm . . . smells like libssh. I wonder if it's related to another problem we are seeing in Linux with the newer clients; we are constantly dropping SSH sessions. Sound, file sharing, printing, all those functions which rely upon the reverse ssh forwarding all break after a while. Suspending and resuming the session restores them - John snip x2go client 3.99.0.1 under ubuntu (with new libssl) works, but not under windows. snip It works for us too under Squeeze in the sense the session is established and all works initially. However, after a while, the sound stops. Does this happen to you? - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] x2goserver goes pluggable
On Thu, 2011-09-29 at 23:20 +0200, Mike Gabriel wrote: Hi all, last night I have implemented a very basic pluggable extension system for x2goserver. The relevant folder structure will be installed by the x2goserver-extensions package. What the package provides is described on the x2goserver-run-extensions man page: http://code.x2go.org/doc/x2goserver-extensions/man/man8/x2goserver-run-extensions.8.html Feedback+Comments+Criticism are Welcome!!! snip I like this idea very much. Thank you - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] x2goserver-printing package
On Sat, 2011-09-24 at 10:47 +0200, Arnold Krille wrote: On Saturday 24 September 2011 05:46:39 John A. Sullivan III wrote: On Sat, 2011-09-24 at 01:53 +0200, Mike Gabriel wrote: Hi all, in a phone conference Heinz, Alex and I have talked about printing not being a very core feature of X2go Server. Esp. the x2goprint use is not needed on X2go Servers that are used without the print facility. Because of this I have made a proposal in X2go Git that move X2go client-side printing support back into a separate package (x2goserver-printing). Please take a look at the committed changes (x2goserver.git) and comment. snip I would agree that is makes sense to keep it as a separate feature. I do hope there is no discussion of removing support as I do think X2Go would be woefully incomplete without it. Thanks - John That is debatable:-) We use X2Go totally without the need for client-side printing. What is the purpose of central printers and central print-servers and diskless thin-clients, when every user gets his own printer and then needs printing at his place. snip grin when there are no central printers and central print-servers or when the client is accessing their desktop from some atypical location and needs to utilize the facilities at hand. So, as you did mention, there are other scenarios - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] x2goserver-printing package
On Sat, 2011-09-24 at 01:53 +0200, Mike Gabriel wrote: Hi all, in a phone conference Heinz, Alex and I have talked about printing not being a very core feature of X2go Server. Esp. the x2goprint use is not needed on X2go Servers that are used without the print facility. Because of this I have made a proposal in X2go Git that move X2go client-side printing support back into a separate package (x2goserver-printing). Please take a look at the committed changes (x2goserver.git) and comment. snip I would agree that is makes sense to keep it as a separate feature. I do hope there is no discussion of removing support as I do think X2Go would be woefully incomplete without it. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] X2Go printing crash [was: libssh, nagle, and performance issues]
On Thu, 2011-09-22 at 07:51 +0200, Oleksandr Shneyder wrote: Am 21.09.2011 22:55, schrieb John A. Sullivan III: I tried to intercept it with ddd but that didn't work. Interestingly, its messages were that the program closed normally as if x2goprinting is invoking a close event. I'll see what I can do to get this to run from a debugger but it is a showstopping bug for us - John snip Hallo John, I still thinking, that client crashes because it cannot work with output from your server. Can you give me a X2Go account on one of your servers, where you have troubles with printing? regards snip Thanks, Alex. I'm working on getting that set up for you. Interestingly, even if I cancel the print job from the x2go print dialog, x2goclient still exits - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] libssh, nagle, and performance issues
On Mon, 2011-09-19 at 14:54 +0200, Mike Gabriel wrote: Hi Alex, On Mo 19 Sep 2011 13:44:15 CEST Oleksandr Shneyder wrote: hmmm, ok, seems you are right :) Can you apply this patch to source tree and I'll try to build a windows binary tomorrow. regards, Alex patch has been committed. I'll build new packages for Debian tonight (I am at a customer location right now). Thanks, Mike snip I can confirm the significant improvement in performance. For example, and what really made me think Nagle, my ping results display much better. When doing ping tests previously, the results would stutter even though the response times were nearly even. In fact, sometimes I would seem to miss a packet only to have two display the next second. They are now perfectly smooth. However, we still have a BIG problem. X2Go client printing still segfaults the entire client every single time we try to print. Phil has confirmed this independently from me. Any ETA on that fix? Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] libssh, nagle, and performance issues
On Mon, 2011-09-19 at 13:44 +0200, Oleksandr Shneyder wrote: Am 19.09.2011 13:30, schrieb Mike Gabriel: Hi Alex, On Mo 19 Sep 2011 09:51:16 CEST Oleksandr Shneyder wrote: Am 17.09.2011 03:39, schrieb John A. Sullivan III: On Fri, 2011-09-16 at 23:17 +0200, Mike Gabriel wrote: Hi John, hi Alex, On Fr 16 Sep 2011 21:56:38 CEST John A. Sullivan III wrote: Hello, all. Since moving to the newer clients using libssh, I've noticed what appears to be a performance degradation. It smells like Nagle coalescing packets, e.g., most of the screen will paint, then it will pause a moment, and then the rest will paint. Is there any chance we have neglected to disable Nagle when invoking libssh? This appears to have affected the SNMP over SSH project: http://books.google.com/books?id=S62is2vVoVgCpg=PA29lpg=PA29dq=libssh +naglesource=blots=F8Iw-pvQxdsig=f5dfy0K0ewzSZOhqzTJRSKsDKZ8hl=enei=KppzTrCZD4LJgQeMxLzzDAsa=Xoi=book_resultct=resultresnum=1sqi=2ved=0CBwQ6AEwAA#v=onepageq=libssh%20naglef=false This is in section 4.1. The difference was a latency of over 800ms to one of under 60ms. That seems to be about what I am seeing. There is a delay of roughly a second and then the rest of the screen comes through. Thanks - John Indeed the TCP_NODELAY (which means: turn Nagle off) is not set for the ssh tunnel in X2goClient. It probably is for the SSH connection itself, as the SNMP patch has been reported to libssh upstream and their patch suggestion has been sensible. However, for the SSH tunnels built up for X2go this TCP_NODELAY option does perhaps not get set. @Alex: would this patch fix that in x2goclient??? diff --git a/sshprocess.cpp b/sshprocess.cpp index 9e72e8f..03ee9b6 100644 --- a/sshprocess.cpp +++ b/sshprocess.cpp @@ -88,6 +88,7 @@ void SshProcess::tunnelLoop() const char y=1; #endif setsockopt(serverSocket, SOL_SOCKET, SO_REUSEADDR,y, sizeof(int)); +setsockopt(serverSocket, IPPROTO_TCP, TCP_NODELAY,y, sizeof(int)); address.sin_family=AF_INET; address.sin_addr.s_addr=INADDR_ANY; I don't think that that can be helpful. This part of code is only for reverse tunnel which is used for FS forwarding and Pulse Audio. I took another close look at the code of sshProcess.cpp. The patch is applied to the method sshProcess::tunnelLoop. This method only gets execute in sshProcess::startTunnel. The call is: quote if (!reverse) tunnelLoop(); else { blablabla } /quote So this makes me think that tunnelLoop() is not for reverse tunneling but for the graphics tunnel. Could you take another look please? Thanks, Mike hmmm, ok, seems you are right :) Can you apply this patch to source tree and I'll try to build a windows binary tomorrow. regards, Alex snip Hi, Alex. Are the Windows packages available anywhere to test? Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] libssh, nagle, and performance issues
On Tue, 2011-09-20 at 16:42 +0200, Oleksandr Shneyder wrote: Am 20.09.2011 12:37, schrieb John A. Sullivan III: On Mon, 2011-09-19 at 14:54 +0200, Mike Gabriel wrote: snip I can confirm the significant improvement in performance. For example, and what really made me think Nagle, my ping results display much better. When doing ping tests previously, the results would stutter even though the response times were nearly even. In fact, sometimes I would seem to miss a packet only to have two display the next second. They are now perfectly smooth. However, we still have a BIG problem. X2Go client printing still segfaults the entire client every single time we try to print. Phil has confirmed this independently from me. Any ETA on that fix? Thanks - John I cannot reproduce it. I've tried to print with win xp and win7. No issues. Can it be an issue related to your special environment? regards, snip I doubt it as the problem is one the client side and not the server side. All our changes were to the server. This is in the Squeeze client and not Windows - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Status load balancing in X2go, and alternatives
On Mon, 2011-09-19 at 22:38 +0200, Paul van der Vlis wrote: Op 16-09-11 07:31, John A. Sullivan III schreef: We are building that scale of environment. For better or worse, we have been so overrun with consulting work that we are way behind where we hoped to be in our X2Go build but we have taken a different approach. We integrated X2Go and Linux-VServer. Each user has a dedicated system (great for things like non-repudiation) but the extra resource requirements are minimal. Even with RAM? I think every user needs his own OpenOffice in RAM, so you need much more RAM this way. But maybe I'm wrong. Nope - with the impressive hashification feature of VServer there is only a single copy of the binary in memory. So if I have four hundred identical systems, I consume the disk space of one system and the memory footprint of one system (not quite realistic as it will have more binaries loaded than a single user would likely use but 300 users may have a single copy of OO in memory, 275 may have a single copy of Firefox, 50 a single copy of GIMP, etc., and the user data may consume considerably more than the binary (especially with massive Firefox memory leaks as we sometimes see). We have been thinking about using Linux containers. But we decide first to work with some shared virtual machines because we have more experience with that. http://lxc.sourceforge.net/ We then moved the x2gocleansession processing to the VServer host so there is only one daemon running every five seconds instead of hundreds. We also adapted the database to use schemas - one per user, and triggers for coordinating individual user changes with the master database used by x2gocleansession and other reporting utilities we created. Thus, we have a single database for all those potentially hundreds of X2Go servers. Finally, we adapted the x2go printing so that we could use a single CUPS server for all X2Go users and not have hundreds of instances of CUPS running. So far it has scaled very nicely but I am way behind in seeing how we would make the same changes to 3.99 - John Does every user have it's own IP on the server? Each user has their own VM and thus their own IP and, since the VMs do not have root capability or system level privileges, they cannot change their IPs (or do things like sniff packets). This works really nicely when we overlay Firepiping on the system (http://iscs.sourceforge.net) so that we can granularly restrict what packets users can send on the network to shut down any snooping or attempts at escalating privileges - John With regards, Paul. ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
[X2go-Dev] libssh, nagle, and performance issues
Hello, all. Since moving to the newer clients using libssh, I've noticed what appears to be a performance degradation. It smells like Nagle coalescing packets, e.g., most of the screen will paint, then it will pause a moment, and then the rest will paint. Is there any chance we have neglected to disable Nagle when invoking libssh? This appears to have affected the SNMP over SSH project: http://books.google.com/books?id=S62is2vVoVgCpg=PA29lpg=PA29dq=libssh +naglesource=blots=F8Iw-pvQxdsig=f5dfy0K0ewzSZOhqzTJRSKsDKZ8hl=enei=KppzTrCZD4LJgQeMxLzzDAsa=Xoi=book_resultct=resultresnum=1sqi=2ved=0CBwQ6AEwAA#v=onepageq=libssh%20naglef=false This is in section 4.1. The difference was a latency of over 800ms to one of under 60ms. That seems to be about what I am seeing. There is a delay of roughly a second and then the rest of the screen comes through. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] Greetings and a lot of questions.
On Thu, 2011-09-15 at 12:21 -0500, Daniel Quiroga wrote: Hi! I am working on a X2go client for Android/iOS snip Wow! I'd love to see that. Good luck - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] X2go Server and syslogging
On Fri, 2011-09-09 at 18:54 +0200, Mike Gabriel wrote: Hi John, On Fr 09 Sep 2011 18:52:22 CEST John A. Sullivan III wrote: On Fri, 2011-09-09 at 18:31 +0200, Mike Gabriel wrote: hi all, today I have added syslog support to all X2go Server scripts. Currently, two log levels are used: debug - mostly for db debugging info - write to syslog whenever one of the X2go Server commands is called The default loglevel is ,,notice''. For this level, I plan to log if a session has been started/suspended/resumed/terminated. More to come later, Mike snip Not that it's critically important but, if you insert the word ERROR when logging an error condition, I believe it can be picked up by OSSEC for log reporting - John Not sure what your request is: shall I avoid that or include that as a feature? Mike Hi, Mike. It was just information. As you craft the message for error conditions, if you use the string ERROR, OSSEC users will be able to pick up those error automatically and alert on them. It would be nice feature for OSSEC users (like us!). Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] x2goserver release tag 3.99.0.6?
On Thu, 2011-08-18 at 08:27 +0200, Mike Gabriel wrote: Hi devs, there are quite some patches coming in for x2goserver (x2goprint, x2goruncommand, etc.). I would like to make a nano release on the x2goserver (3.99.0.6), tag it in Git (no Git version of the server has yet been tagged) and also push the package towards the stable repositories (main area for packages.x2go.org, stable ppa on LP). Comments? snip Could you give a quick summary of the changes? Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Huge browsing delays
On Tue, 2011-08-16 at 18:16 +0200, Moritz Struebe wrote: On 2011-08-15 19:55, John A. Sullivan III wrote: We can produce this without fail in our Zimbra environment. There is an Zimbra live demo at http://www.zimbra.com/products/hosted_demo.php I would think you could run this from any X2Go server although the reliable problem for us is chat and I do not know if the Zimbra demo has chat. I just tested FF6. To me it seems to much more responsive on js-intensive sites (in my case google maps). Could you give that a try? I know it's poking in the dark, but to me the whole issue seems odd as it was not yet possible to put a finger onto the reason. (Probably it's some bug in the server that gets hidden by most applications or something on the client doing something odd ;) ) Hmm.. . I'm not sure FF6 will be an option for us. I'll have a look and see what I can do. I don't know if that is an option for John Williams. On the other hand, I don't think we can say to folks X2Go works only for users with FF6 or newer :( Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] @John S., Alexander C.: wiki access
On Tue, 2011-08-16 at 19:37 +0200, Mike Gabriel wrote: Hi all, esp. John Sullivan and Alexander Crössmann, the X2go roadmap states that you two guys would have liked to write some wiki documentation. At least the wiki roadmap says so. right at the top of the page... http://wiki.x2go.org/roadmap Do you already have access? Can I remove that line? (you could that yourself...). snip I am quite keen to do so but, right now, I'm up to my eyeballs in work - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Huge browsing delays
On Sun, 2011-08-14 at 18:44 +0200, Mike Gabriel wrote: Hi John, hi all, On Mi 03 Aug 2011 16:18:48 CEST John A. Sullivan III wrote: 3.0.1-4 Windows client to 3.99.0.0 server with X2GoAgent 3.5: No delay whatsoever - and that's after pounding it for a long time. This basically _very_ good new as it tells us that we have to look on the client side (I hope). I know nothing about NX internals so I am just speculating here. The problem only seems to be with heavy AJAX applications as far as I can tell. Could it be related to how AJAX only paints part of the screen? Could it be that NX does not pick up the screen change because if is only a subportion of the application Window and done in a way different from, for example, OpenOffice updates the area being typed? Could it be that every certain number of seconds, NX checks the entire screen for updates and that this timer is different in the various configurations and maybe non existent with two mismatched Agents? So we should list up the differences between x2goclient 3.0.1-4 (Windows and 3.99.0.x (also on Windows). There are many... Who can do this? I have no idea but I would consider this a blocking bug. I'll try to squeeze some xkbwatch in this week - John Can you, John, or anyone else give a howto that explains how to reproduce this problem on a fresh server install (Debian squeeze) and a fresh client install (Windows, what version?, what arch? what Xserver? Etc.). We have to try to address this issue in a very generic way if we/you want it solved. Expecting a debugging recipe to trace this bug... THANKS! We can produce this without fail in our Zimbra environment. There is an Zimbra live demo at http://www.zimbra.com/products/hosted_demo.php I would think you could run this from any X2Go server although the reliable problem for us is chat and I do not know if the Zimbra demo has chat. I think John Williams and others had readily reproducible non-Zimbra procedures to produce this error including GMail (I do not know which functions). As I mentioned. the problem should be reproducible from any X2Go server but, if you'd like, I can give you the credentials to login to one of our demo systems. Let me know if you prefer to go that route. This and the printing crash are the two blocking bugs in my assessment. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Mon, 2011-08-01 at 22:58 +0200, Mike Gabriel wrote: Hi John, On Mo 01 Aug 2011 20:43:23 CEST John A. Sullivan III wrote: Out of curiosity, I tried reproducing this manually. Both systems where I can reliably reproduce this problem have a local smtp daemon. I did a ssh -p some hidden port -L 40015:localhost:25 user1@machine1 and a ssh -p some hidden port -L 40016:localhost:25 user2@machine2. I was able to telnet to each on port 40015 and 40016 without an issue or one disconnecting the other. Could there be a problem in the way libssh is handling multiple channels? - John BRILLIANT, that is it!!! Debian squeeze still uses libssh-4 0.4.5, but x2goclient needs a patch (sent to libssh upstream by Oleksandr) that has entered libssh-4 with version 0.4.7. I have already uploaded a Debian squeeze version of libssh-4 0.4.7. I am also re-building x2goclient against libssh-4 0.4.7. Sorry, I presumed that the newer libs were already in Debian squeeze... Packages will be available soon on packages.x2go.org. snip Hi, Mike. I have libssh-4.0.4.7 on my client: ii libssh-4 0.4.7-1.1 A tiny C SSH library Yet I still have this problem. I cannot connect to more than one X2Go Server. Is there something else I need to do? Client is: ii x2goclient 3.99.0.0-0~x2go2+squeeze~heuler~20110 Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Huge browsing delays
On Tue, 2011-08-02 at 14:21 -0400, John A. Sullivan III wrote: On Tue, 2011-08-02 at 11:15 -0700, John Williams wrote: On Tue, Aug 2, 2011 at 11:09 AM, John A. Sullivan III jsulli...@opensourcedevel.com wrote: I wondered if it was because our servers are still using 3.0.1-5 as this smells like it could be an NXAgent problem. I connected to an X2Go server using the latest X2GoAgent (3.5.0.2-0~x2go1 +squeeze~heuler~201107) and experienced the same problem. I have backleveled my Windows system to 3.0.1-4 to see if the problem goes away but do not have enough data yet to report. Thanks - John I am also using server 3.0.1-5. Here is a summary of whether I have seen this bug with different windows client versions (64-bit Windows 7 OS) on server 3.01-5: windows client version bug occurs? == 3.01-13NO 3.01-18YES 3.99 YES snip I should add I am testing from Windows XP - John snip I was able to gain some interesting, confounding, and possibly important data on this problem today. I still haven't hooked up xkbwatch but I did try a number of different combinations of clients and servers with astoundingly different results. The above results posted by John Williams, my results on a XP platform, and Phil's results with 3.0.18 in Linux all have the 16-20 second delay when speaking to either 3.0.1-5 or 3.99.0.0 servers with both X2GoAgents 3.4 and 3.5. However, I tried two new combinations today: 3.0.1-4 Windows client to 3.0.1-5 server: frequent delay but 3 - 4 seconds - much, much shorter. This explains why our clients are upset but not throwing us out the door as they are all using this combination. 3.0.1-4 Windows client to 3.99.0.0 server with X2GoAgent 3.5: No delay whatsoever - and that's after pounding it for a long time. I know nothing about NX internals so I am just speculating here. The problem only seems to be with heavy AJAX applications as far as I can tell. Could it be related to how AJAX only paints part of the screen? Could it be that NX does not pick up the screen change because if is only a subportion of the application Window and done in a way different from, for example, OpenOffice updates the area being typed? Could it be that every certain number of seconds, NX checks the entire screen for updates and that this timer is different in the various configurations and maybe non existent with two mismatched Agents? I have no idea but I would consider this a blocking bug. I'll try to squeeze some xkbwatch in this week - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Tue, 2011-08-02 at 09:50 +0200, Moritz Struebe wrote: On 2011-08-02 01:34, John Williams wrote: I concur. This is the most important bug as far as I am concerned. It is stopping me from running any windows client after 3.01-13 (except for testing purposes). I've been running 3.01-13 for months now in 64-bit Windows 7 with only a few bugs that are merely annoying, but I have never been able to run anything after 3.01-13 due to this bug which is a show stopper. Have you looked into the issues using rdp? There are several network tweaks that are supposed to help - they might help here, too. snip Yes, I appreciated the lead and spent a few hours playing with tweaking the autotuning parameters even though the description didn't seem to quite apply - those problems seemed more generic. Alas, it did not fix the problem - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] rename of x2godesktopsharing group
On Tue, 2011-08-02 at 16:59 +0200, Mike Gabriel wrote: Hi, On Di 02 Aug 2011 16:46:56 CEST Reinhard Tartler wrote: On Tue, Aug 02, 2011 at 15:41:55 (CEST), Mike Gabriel wrote: Hi Morty, On Di 02 Aug 2011 15:28:57 CEST Moritz Struebe wrote: What is the rationale for the extra group? Is there a good reason for disallowing someone to share his/her desktop? the desktopsharing is a tricky feature anyway as it grants many ways for a user who is allowed to share another's desktop to manipulate the user profile. This sounds to me as desktop sharing was a somewhat insecure feature anyway. In this case, why do you rely on a system group instead of for instance maintaining a /etc/x2go/allowdesktopshareing.users file that contains all users that are allowed to use the feature? x2godesktopsharing falls into a daemon (in user space) and a client (the systray)--I think it is this way around. And these two communicate via a socket file. And the write access is granted by group membership. No group membership, no desktop access. Greets, Mike snip I haven't thought it through thoroughly but I believe what you propose makes sense. Allowing it to be disabled allows one to shut down the social engineering vector. Yes, users need to grant access but they also do when they should not - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Mon, 2011-08-01 at 22:58 +0200, Mike Gabriel wrote: Hi John, On Mo 01 Aug 2011 20:43:23 CEST John A. Sullivan III wrote: Out of curiosity, I tried reproducing this manually. Both systems where I can reliably reproduce this problem have a local smtp daemon. I did a ssh -p some hidden port -L 40015:localhost:25 user1@machine1 and a ssh -p some hidden port -L 40016:localhost:25 user2@machine2. I was able to telnet to each on port 40015 and 40016 without an issue or one disconnecting the other. Could there be a problem in the way libssh is handling multiple channels? - John BRILLIANT, that is it!!! Debian squeeze still uses libssh-4 0.4.5, but x2goclient needs a patch (sent to libssh upstream by Oleksandr) that has entered libssh-4 with version 0.4.7. I have already uploaded a Debian squeeze version of libssh-4 0.4.7. I am also re-building x2goclient against libssh-4 0.4.7. Sorry, I presumed that the newer libs were already in Debian squeeze... Packages will be available soon on packages.x2go.org. snip So if we are running older versions of X2Go server, what do we update? I assume the libssh on the client. Does this imply there will be a new Windows client? It looks like it is still 3.99 on the wiki. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
[X2go-Dev] 3.99 Testing - Huge browsing delays
Hello, all. I thought I would spawn this subject into a separate thread as it is a very serious issue. To summarize, when using firefox/iceweasel in X2Go, the browser frequently hangs for between 16 and 20 seconds. All other applications work perfectly fine, accept input, and display output, but the browser window remains frozen. Once it unfreezes, all input and output is processed, e.g., typed letters are displayed and tab clicks are executed. It appears to happen in AJAX intensive sites (Zimbra, GMail, GLPI) but this is not definitive. We initially thought it was a Windows only problem but Phil has confirmed he experiences this regularly in Linux using client 3.0.1.18. I confirm it on Windows 3.99. I wondered if it was because our servers are still using 3.0.1-5 as this smells like it could be an NXAgent problem. I connected to an X2Go server using the latest X2GoAgent (3.5.0.2-0~x2go1 +squeeze~heuler~201107) and experienced the same problem. I have backleveled my Windows system to 3.0.1-4 to see if the problem goes away but do not have enough data yet to report. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Huge browsing delays
On Tue, 2011-08-02 at 12:49 -0700, John Williams wrote: On Tue, Aug 2, 2011 at 11:39 AM, Moritz Struebe moritz.stru...@informatik.uni-erlangen.de wrote: The smells like Nagle, What is Nagle? snip Nagle is an algorithm which coalesces small TCP responses into larger ones to improve throughput at the expense of latency. In X2Go, we generally want the lowest latency possible - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Mon, 2011-08-01 at 13:30 -0400, John A. Sullivan III wrote: On Fri, 2011-07-29 at 10:21 +0200, Mike Gabriel wrote: Hi Daniel, On Fr 29 Jul 2011 07:24:40 CEST Daniel Lindgren wrote: [...] I've been trying to think of some way to make x2go automatically only use free ports, but since there are possibly several clients and servers involved in multiple incoming and outgoing sessions, the only safe way I've come up with would be to register used ports (or reserve port ranges) on a shared resource somewhere, which would make all x2go clients and servers dependent on that one resource. What you have been describing above is the old, well known NXServer/NXClient problem. To my experience it does not exist with X2go (at least not with pyhoca-gui ;-) ). The connecting stuff is a bit tricky... - x2goclient connects via SSH and launches either x2gostartagent or x2goresume-session - x2gostartagent will try to detect a free server-side port 3. This port is put into the session database table. (Make sure you do not have other server-side stuff running that claims ports between 3 and 3++. - x2goresume-session will pick the formerly detected port from the session database table (i.e. SQLite or postgres) and presume it is still unused (as it should be, unless unused by X2go - x2gostartagent avoids claiming ports that are in the database marked with S (suspended) or R (running) state - say we detect a free server-side port: 30015: x2gostartagent will launch the XNest-like x2goagent XServer, listening for incoming connections on port localhost:300015. - x2goclient now tries to set up a forwarding tunnel from client to server (-L 3:localhost:30015) - if the local port 3 (the first one in the above expression) is already in use, it simply selects another port. Each selected port is probed before usage. That is done by x2goclient. - say we end up with -L 30028:localhost:30015 The problem you observe: ”the remote proxy closed the connection while negotiating …” to my point of view stems from uncleanly closed SSH connections. Try to restart the SSH daemon when this error occurs and you should be able to connect again. This error can also be caused by the sound/sshfs tunnels. Here the same cause applies: unclean shutdown of SSH connections. Reasons may be: o unclean x2goclient/pyhoca-gui code o line failures (so the clients do not get the chance to cancel the port forwarding requests...) You may do the following to debug: o try if the problem occurs in the same way when you use pyhoca-gui/-cli o pyhoca-gui (i.e. python-x2go) tries to restore the mal-closure of SSH port forwardings... You may try to resume a session twice, the second time you should get a session window The addressed issue was a big problem for Python X2go before version 0.1.0.0. It took me quite some sleepness nights to get rid of it by 90%+. snip I continue to have a devil of a time with this. It seems like - x2goclient now tries to set up a forwarding tunnel from client to server (-L 3:localhost:30015) - if the local port 3 (the first one in the above expression) is already in use, it simply selects another port. Each selected port is probed before usage. That is done by x2goclient. isn't working. In the past, using older clients, we would generate error messages about a problem connecting on port 30001 but the sessions would all still work. We are no longer seeing those errors. Instead, the sessions are rudely dropped. I've restarted all the X2Go servers and clients after deleting all the old session directories on both. Still no success. Here are the details. In our environment, we have a single PostgreSQL database for all the X2Go servers. There is one X2Go server for each user (x2go client). I connect to the first server and I start a valid session. I am using ports 30001 - 30003 on the server side and 30004 on the client side: tcp0 0 0.0.0.0:30004 0.0.0.0:* LISTEN Server User Status Last Connection Run Time Display Client GP SP FP Context Load RSS VSZ jasiii jasiiiS 01.08.11 12:45:12 0d 0h 2m 51 74.75.231.235 30001 30002 3000340016 0.04 434.2Mb 8.4Gb I start a second x2goclient to connect to the second server. The client is listening on port 30005: tcp0 0 0.0.0.0:30004 0.0.0.0:* LISTEN tcp0 0 0.0.0.0:30005 0.0.0.0:* LISTEN The server side is using the same ports as the other x2go server: Server User
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Mon, 2011-08-01 at 14:38 -0400, John A. Sullivan III wrote: On Mon, 2011-08-01 at 13:30 -0400, John A. Sullivan III wrote: On Fri, 2011-07-29 at 10:21 +0200, Mike Gabriel wrote: Hi Daniel, On Fr 29 Jul 2011 07:24:40 CEST Daniel Lindgren wrote: [...] I've been trying to think of some way to make x2go automatically only use free ports, but since there are possibly several clients and servers involved in multiple incoming and outgoing sessions, the only safe way I've come up with would be to register used ports (or reserve port ranges) on a shared resource somewhere, which would make all x2go clients and servers dependent on that one resource. What you have been describing above is the old, well known NXServer/NXClient problem. To my experience it does not exist with X2go (at least not with pyhoca-gui ;-) ). The connecting stuff is a bit tricky... - x2goclient connects via SSH and launches either x2gostartagent or x2goresume-session - x2gostartagent will try to detect a free server-side port 3. This port is put into the session database table. (Make sure you do not have other server-side stuff running that claims ports between 3 and 3++. - x2goresume-session will pick the formerly detected port from the session database table (i.e. SQLite or postgres) and presume it is still unused (as it should be, unless unused by X2go - x2gostartagent avoids claiming ports that are in the database marked with S (suspended) or R (running) state - say we detect a free server-side port: 30015: x2gostartagent will launch the XNest-like x2goagent XServer, listening for incoming connections on port localhost:300015. - x2goclient now tries to set up a forwarding tunnel from client to server (-L 3:localhost:30015) - if the local port 3 (the first one in the above expression) is already in use, it simply selects another port. Each selected port is probed before usage. That is done by x2goclient. - say we end up with -L 30028:localhost:30015 The problem you observe: ”the remote proxy closed the connection while negotiating …” to my point of view stems from uncleanly closed SSH connections. Try to restart the SSH daemon when this error occurs and you should be able to connect again. This error can also be caused by the sound/sshfs tunnels. Here the same cause applies: unclean shutdown of SSH connections. Reasons may be: o unclean x2goclient/pyhoca-gui code o line failures (so the clients do not get the chance to cancel the port forwarding requests...) You may do the following to debug: o try if the problem occurs in the same way when you use pyhoca-gui/-cli o pyhoca-gui (i.e. python-x2go) tries to restore the mal-closure of SSH port forwardings... You may try to resume a session twice, the second time you should get a session window The addressed issue was a big problem for Python X2go before version 0.1.0.0. It took me quite some sleepness nights to get rid of it by 90%+. snip I continue to have a devil of a time with this. It seems like - x2goclient now tries to set up a forwarding tunnel from client to server (-L 3:localhost:30015) - if the local port 3 (the first one in the above expression) is already in use, it simply selects another port. Each selected port is probed before usage. That is done by x2goclient. isn't working. In the past, using older clients, we would generate error messages about a problem connecting on port 30001 but the sessions would all still work. We are no longer seeing those errors. Instead, the sessions are rudely dropped. I've restarted all the X2Go servers and clients after deleting all the old session directories on both. Still no success. Here are the details. In our environment, we have a single PostgreSQL database for all the X2Go servers. There is one X2Go server for each user (x2go client). I connect to the first server and I start a valid session. I am using ports 30001 - 30003 on the server side and 30004 on the client side: tcp0 0 0.0.0.0:30004 0.0.0.0:* LISTEN Server User Status Last Connection Run Time Display Client GP SP FP Context Load RSS VSZ jasiii jasiiiS 01.08.11 12:45:12 0d 0h 2m 51 74.75.231.235 30001 30002 3000340016 0.04 434.2Mb8.4Gb I start a second x2goclient to connect to the second server. The client is listening on port 30005: tcp0 0 0.0.0.0:30004 0.0.0.0
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Thu, 2011-07-28 at 07:38 -0400, John A. Sullivan III wrote: On Thu, 2011-07-28 at 08:59 +0200, Oleksandr Shneyder wrote: Am 28.07.2011 08:51, schrieb John Williams: On Wed, Jul 27, 2011 at 11:46 PM, Oleksandr Shneyder oleksandr.shney...@obviously-nice.de wrote: If that is the same issue as the freezes (or extremely slow response) that I see when I open a gmail tab in firefox, then the issue is completely absent with Windows x2goclient 3.01-13. ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev I don't think, that is a client problem. X2Go client 3.01-13 using the same version of nxproxy as the 3.01-19. As I understood this is not a specific windows problem? this problem exists although with linux client? snip Windows only for me. Thanks - John snip In fact, this one needs some serious attention. I'm back in a Windows world today and I'm amazed at how patient our clients have been. This is unusable. One can't be expected to wait 20 seconds between mouse clicks or keyboard strokes but that is what is happening when this bug kicks in. What should take twenty seconds literally takes five minutes! Again, this is only in Windows and only when using X2Go - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] PROPOSAL: X2go session cache deletion (client-side)
On Fri, 2011-07-29 at 11:04 +0200, Mike Gabriel wrote: Hi Morty, On Fr 29 Jul 2011 10:55:27 CEST Moritz Struebe wrote: On 2011-07-29 10:39, Mike Gabriel wrote: 1. Place file ,,profilename'' in the .x2go/session-name directory to become able to relate between server and client. Why is this needed? I do think we must relate a session profile to a server, though. Especially when using a shared home, but a local sqlite database. Currently we do not know which .x2go/session dir belongs to what session profile (=server) if the server-side session has terminated. That is: if the server-side session has vanished, we do not know if an .x2go/session is actually a terminated session on session profile A or a still running session on session profile B. We could also look for entries in session.log that inform about a terminated session: mike@mimino:~/.x2go$ grep -ri session terminated * S-locadm-50-1311057154_stDGNOME_dp24/session.log:Session: Session terminated at 'Tue Jul 19 08:33:13 2011'. S-locadm-51-1311437782_stDGNOME_dp24/session.log:Session: Session terminated at 'Sat Jul 23 16:49:39 2011'. S-locadm-52-1309380386_stDGNOME_dp24/session.log:Session: Session terminated at 'Wed Jun 29 21:42:50 2011'. S-locadm-52-1311438419_stDGNOME_dp24/session.log:Session: Session terminated at 'Sat Jul 23 17:06:20 2011'. S-locadm-61-1311281243_stDGNOME_dp24/session.log:Session: Session terminated at 'Sat Jul 23 16:43:56 2011'. S-locadm-61-1311281243_stDGNOME_dp24/session.log:Session: Session terminated at 'Sat Jul 23 17:14:09 2011'. However, what I do not like about this last approach is: it is NX specific. I would rather like some x2go specific method to grip here... snip I would also like to see it NX independent. I could see us integrating SPICE within a year - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Fri, 2011-07-29 at 07:24 +0200, Daniel Lindgren wrote: 2011/7/28 John A. Sullivan III jsulli...@opensourcedevel.com: snip I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John One thing I've had problems with in the past is TCP port conflicts. When x2go starts a session, it always uses a TCP port above 3, decided by x2gostartagent on the server side. If you start multiple sessions from one client to several different servers, wouldn't that create a port conflict sooner or later? Maybe that is what you've seen, with conflicts causing sessions to terminate. Example, x2go on client CA connects (a ssh tunnel) to server SA on port 30001 (-L 30001:localhost:30001 user@SA). You then start another session from client A to server B, also using port 30001 (-L 30001:localhost:30001 user@SB). Doesn't that create a port conflict on client CA? You have two separate ssh tunnels trying to use the same TCP port on the client. The problems I've seen have been when client CA has connected to server SA, which then has been used as a gateway client to connect to other servers. Sometimes the session between CA and SA used port 30001 and the session(s) between SA and other server SB also used port 30001. I could always set up the sessions, but if I tried to reconnect from client CA to SA I got an error message about ”the remote proxy closed the connection while negotiating …”. My workaround was to separate the port ranges used by x2gostartagent on the servers; server SA now uses 30001-, server SB uses 30101-, server SC uses 30201- and so on. No more conflicts. I've been trying to think of some way to make x2go automatically only use free ports, but since there are possibly several clients and servers involved in multiple incoming and outgoing sessions, the only safe way I've come up with would be to register used ports (or reserve port ranges) on a shared resource somewhere, which would make all x2go clients and servers dependent on that one resource. snip Yes, but I believe in the past it has always managed to magically pick another set of ports. That seems to have stopped happening - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Wed, 2011-07-27 at 12:37 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 12:18 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 09:33 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 14:41 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 14:33, schrieb Mike Gabriel: Hi Alex, On Mi 27 Jul 2011 13:52:46 CEST Oleksandr Shneyder wrote: Am 22.07.2011 06:38, schrieb John A. Sullivan III: I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hello John, Unfortunately, I can't reproduce this. I was able to start a several sessions simultaneously on debian and win7, see screen shots: http://www.x2go.org/screenshots/multisess-debian.png http://www.x2go.org/screenshots/multisess-win7.png Thanks for all your testing If I recall John correctly, her meant: start x2goclient, connect to session A as user A start another x2goclient, also connect to session A as user A (that is a second session as the same user to the same server) John, could you please cross-check this? Thanks, Mike As I understood, John meant, connecting with user A to server A, start second client and connecting as user B to server B. But it does not matter, I've tried all variants: A-A B-B A-A B-A A-A A-B A-A A-A regards, alex snip Oleksandr is correct - two separate users connected to two separate X2Go servers. I can do the initial connection but, within a few seconds, the first session times out. I've seen this myself on Debian and Windows XP and another tester confirmed it on Windows 7 64 bit. I'll try to dig out more details. Thanks very much for investigating - John snip OK - now this is bizarre - after days of it not working on Squeeze, it decided to work today when I tested. All right, who changed things while I wasn't looking! I'll give Windows a test later but I do know I wasn't imagining this and we had another tester confirm the same phenomenon. I'll check again with them, too. Thanks - John snip The bizarreness continues. Now it works in Windows, too. Still waiting to hear back from the other tester. Other things have suddenly started to work, too, as I'll report separately, yet we've done nothing to change anything in our environment :( snip I can confirm this is still a problem for some. The below report is from a windows 7/64 user. How do we begin troubleshooting? Thanks - John Same problem - soon after starting the second window, the system comes up with an error message: No response received from the remote server. Do you want to terminate the current session?. If one selects No, one is able to connect and use the current (second) session, but it immediately kills the previously started (first) session, with the message The connection with the remote session was shut down. Please check the state of your network connection. If one then attempts to start another (third) session, it generates the second error message again, which prevents a new session from starting, but does leave the existing (second) sesson running. Not sure if there is a clue here, but when trying to log into the third session (when the session-locking dialog box comes up), I note that the number of black bullets coming up is different from the number of keys that I have pressed - and that if one tries to erase them with the backspace key, each press generates another black bullet. So no progress viewed from this end, I regret. ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Thu, 2011-07-28 at 08:59 +0200, Oleksandr Shneyder wrote: Am 28.07.2011 08:51, schrieb John Williams: On Wed, Jul 27, 2011 at 11:46 PM, Oleksandr Shneyder oleksandr.shney...@obviously-nice.de wrote: If that is the same issue as the freezes (or extremely slow response) that I see when I open a gmail tab in firefox, then the issue is completely absent with Windows x2goclient 3.01-13. ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev I don't think, that is a client problem. X2Go client 3.01-13 using the same version of nxproxy as the 3.01-19. As I understood this is not a specific windows problem? this problem exists although with linux client? snip Windows only for me. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
Hi, Alex. Thanks again for taking this seriously. I'll respond in the text - John On Thu, 2011-07-28 at 08:46 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 18:44, schrieb John A. Sullivan III: Hi, Alex. Although the packing and multi-session problems have mysteriously vaporized as I reported elsewhere, I can confirm that we still have the following: 1) Full screen toggles do not work in Windows - no magic pixel, AltCtl-M, AltCtl-F. 2) AltTab does not work as expected in Full Screen mode - in fact, rather than scroll through the X2Go desktop applications, it is currently the only way to break out of full screen! 1 and 2 are xvcsrv issues. As I can remember, It was your idea to start X in full screen mode by full screen X2Go session. The magic pixel and AltCtl-M, AltCtl-F are possible to use only with window manager (on windows only if vcxsrv started in rootless mode). The only way to do this with vcxsrv in full screen mode is find a WID of vcxsrv and send WS_MINIMIZE to it main window. To make it possible we should modify both - x2goagent and x2goclient. My opinion - it is wrong idea. Vcxsrv should be started in rootless mode even for full screen X2Go sessions. If it is not working correctly, we should communicate with developers of vcxsrv to solve this issue. That makes sense except, if I remember correctly, there is a difference between full screen and rootless. If I choose fullscreen, it truly is full screen and the user perceives only one desktop. It is as if their X2Go desktop is their only desktop and they do not have the problem of frequently closing their desktop accidentally when closing a maximized application, i.e., clicking on the wrong x in the top corner. This setting is highly desirable for those who work only or primarily in their X2Go desktops (the majority of our clients). In rootless, the X2Go desktop is still placed within a Windows but it takes the entire available area aside from the window decorations. This is highly desirable for those who frequently toggle between their physical and x2go desktops. I would think we would want to accommodate both (and more!) styles of work. When in full screen mode, my desire would be that the keystrokes would work as if the x2go desktop was the only desktop, i.e., Alt Tab cycles through the x2go desktop apps. The same is with alt+tab. This events should catch vcxsrv. You can specify by x2goclient the X-server you want to use and the command line arguments to starting it. As far as you find a best combination of vcxsrv version and command line arguments I'll make it default. 3) I get a knotify: no process found error when exiting the Windows session. old workaround for kde3, I'll take care about it. 4) Evolution selection (as a representative graphics issue) is still painfully slow Still can not reproduce it. But even if I can, it is a nxlib problem and will be not easy to solve. All output of nxlibs you can find on client and server side in ~/.x2go folder. Have you this problem only with recent versions of nx or was it always present? If it is a new problem, please find a last version of nxlibs without this problem, it can help to find a bug by researching of changelogs and diffs. I'll have a look and try to dig deeper. It is a Windows only problem (or perhaps I should say Windows exacerbated). As I mentioned earlier, I do have a concern that a number of these issues may not be specifically X2Go but rather vcxsrv or NX. The vcxsrv dev has been fabulously supportive. I don't know about NX or if we have the capability to fix it ourselves. Thanks again - John snip ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - stray deletes
On Thu, 2011-07-28 at 19:59 +0200, Daniel Lindgren wrote: 2011/7/28 John A. Sullivan III jsulli...@opensourcedevel.com: This does not seem to be limited to deletes. I've been doing a lot of presentation work today. I'll have paged down a couple of slides and then will be reading and thinking about the current slide for literally several minutes. Suddenly, the next slide pops up as if the page down has been replayed. Is anyone else seeing this? This is 3.99 on both client and server - both Debian Squeeze - John Yes, I've seen random Page Down too. They can come after a long time of inactivity (minutes). Also using Debian Squeeze och both client and server. snip On Michael's recommendation, I have upgraded to heuler and have not seen this problem yet. I will keep the list apprised. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Thu, 2011-07-28 at 07:29 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 12:37 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 12:18 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 09:33 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 14:41 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 14:33, schrieb Mike Gabriel: Hi Alex, On Mi 27 Jul 2011 13:52:46 CEST Oleksandr Shneyder wrote: Am 22.07.2011 06:38, schrieb John A. Sullivan III: I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hello John, Unfortunately, I can't reproduce this. I was able to start a several sessions simultaneously on debian and win7, see screen shots: http://www.x2go.org/screenshots/multisess-debian.png http://www.x2go.org/screenshots/multisess-win7.png Thanks for all your testing If I recall John correctly, her meant: start x2goclient, connect to session A as user A start another x2goclient, also connect to session A as user A (that is a second session as the same user to the same server) John, could you please cross-check this? Thanks, Mike As I understood, John meant, connecting with user A to server A, start second client and connecting as user B to server B. But it does not matter, I've tried all variants: A-A B-B A-A B-A A-A A-B A-A A-A regards, alex snip Oleksandr is correct - two separate users connected to two separate X2Go servers. I can do the initial connection but, within a few seconds, the first session times out. I've seen this myself on Debian and Windows XP and another tester confirmed it on Windows 7 64 bit. I'll try to dig out more details. Thanks very much for investigating - John snip OK - now this is bizarre - after days of it not working on Squeeze, it decided to work today when I tested. All right, who changed things while I wasn't looking! I'll give Windows a test later but I do know I wasn't imagining this and we had another tester confirm the same phenomenon. I'll check again with them, too. Thanks - John snip The bizarreness continues. Now it works in Windows, too. Still waiting to hear back from the other tester. Other things have suddenly started to work, too, as I'll report separately, yet we've done nothing to change anything in our environment :( snip I can confirm this is still a problem for some. The below report is from a windows 7/64 user. How do we begin troubleshooting? Thanks - John Same problem - soon after starting the second window, the system comes up with an error message: No response received from the remote server. Do you want to terminate the current session?. If one selects No, one is able to connect and use the current (second) session, but it immediately kills the previously started (first) session, with the message The connection with the remote session was shut down. Please check the state of your network connection. If one then attempts to start another (third) session, it generates the second error message again, which prevents a new session from starting, but does leave the existing (second) sesson running. Not sure if there is a clue here, but when trying to log into the third session (when the session-locking dialog box comes up), I note that the number of black bullets coming up is different from the number of keys that I have pressed - and that if one tries to erase them with the backspace key, each press generates another black bullet. So no progress viewed from this end, I regret. snip This user (who happens to be working on an unexpected potentially multi-thousand seat deployment based upon multiple simultaneous sessions per user) confirmed that after upgrading from 3.0.1.14 to 3.99.0.0, he still has the same
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Wed, 2011-07-27 at 14:41 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 14:33, schrieb Mike Gabriel: Hi Alex, On Mi 27 Jul 2011 13:52:46 CEST Oleksandr Shneyder wrote: Am 22.07.2011 06:38, schrieb John A. Sullivan III: I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hello John, Unfortunately, I can't reproduce this. I was able to start a several sessions simultaneously on debian and win7, see screen shots: http://www.x2go.org/screenshots/multisess-debian.png http://www.x2go.org/screenshots/multisess-win7.png Thanks for all your testing If I recall John correctly, her meant: start x2goclient, connect to session A as user A start another x2goclient, also connect to session A as user A (that is a second session as the same user to the same server) John, could you please cross-check this? Thanks, Mike As I understood, John meant, connecting with user A to server A, start second client and connecting as user B to server B. But it does not matter, I've tried all variants: A-A B-B A-A B-A A-A A-B A-A A-A regards, alex snip Oleksandr is correct - two separate users connected to two separate X2Go servers. I can do the initial connection but, within a few seconds, the first session times out. I've seen this myself on Debian and Windows XP and another tester confirmed it on Windows 7 64 bit. I'll try to dig out more details. Thanks very much for investigating - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Wed, 2011-07-27 at 13:47 +0200, Oleksandr Shneyder wrote: snip Thanks, Alex, for taking the time to investigate. I'll comment in the text - John Hello all, last days I spent some time to test the latest version of x2goclient (3.99.0) on different platforms: Server debian lenny i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x debian squeeze amd64, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x Client debian lenny i386 debian squeeze i386 debian squeeze amd64 windows7-32 windows7-64 I've tested connections in LAN and over Internet and was able to connect and work without any serious issues. snip Unfortunately, I can not reproduce some bugs reported of users on this mailing list - Issues with image packaging, Issues with scrolling in firefox In my reporting, these two issues are related. The slow Firefox scrolling is from using pack=none. As I mentioned, the packing problem may not be strictly 3.99 but it is definitely a problem. I have it on Windows XP 32 bit, Windows 7 64 bit, Windows Server 2008 64 bit. I can confirm all three connecting to 3.0.1-5 X2GoServer and Windows XP on both 3.0.1-5 and 3.99.0.0. The XP system was a complete rebuilt from the recovery CD with all the latest patches applied (for hours and hours :( ). It did appear to work correctly with 3.0.1-5 server until I installed PyHoca. That's when the problems started. However, the Windows 7 and Windows Server 2008 systems never had Pyhoca installed. They have a fresh installation of 3.99.0. Oops, 2008 had 3.0.1-4. I just removed it, deleted all related files from Program Files, cleaned out the registry of every reference to x2go and obviouslynice and installed 3.99.0. Same problem immediately. Issues with delays after pressing Up/Down arrow keys. This is not so much an issue with Up/Down arrow keys. I was using it as an example of how dramatically slower X2Go is on Windows. I was using email selection in Evolution on a large Inbox to illustrate the point as Evolution has particularly bad screen handling. snip I can imagine, that some of this bugs are system dependent. X2Go uses many components - x2go server, client, nxlibs, nx(x2go)agent. Under windows x2goclient starts although X-server, SSH server and Pulseaudio server. All this components are depended from many system libraries, for example libpng, libjpeg, etc. Unfortunately, we can not to test X2Go with all possible combinations of software installed on user computer. We running our tests on fresh installed OS with default set of software. And default options of x2goclient (on Windows with default X-server and default command line options) If you expecting troubles running X2Go, please try, if it possible, to install a clean OS, without any additional software (f.e. third-part firewalls or antivirals on Windows) If X2Go running on such clean system, you can compare it with you productive system and find out which software blocking functionality of X2Go. As example, one of our customers found out, that x2go client on Win7 can not connect to server if Avast antivirus is installed. Quite understood and, as I think you know, I'll do whatever I can do to help but we do need to solve these issues. The packing one is critical. It may be a particular combination of packages and I'll enumerate those next as best I can. However, we can tell our clients to reinstall the client but we can't very well tell them to reinstall their OS to ensure it is clean. We'll need to find out what combination causes this so we can both correct it and warn against it. I've already mentioned the client sides and the Windows packaging is pretty self contained. Here's what I see on the server side: 3.0.1-5: ii x2goagent 3.4.0-3-1 Special NX-ized client, was developed by NoMachine to handle X sessions un x2goclientnone (no description available) ii x2gognomebindings 2.0.1-1 Description: x2gognomebindings (GNOME mimetypes for x2go) ii x2gokdebindings 2.0.1-4 x2gokdebindings (KDE mimetypes and icons for x2go) ii x2goprint 3.0.0-1 x2goprint command ii x2goserver3.0.1-5 x2goserver (daemon and tools) ii nxcomp3.4.0-1-1 NX compression library ii nxcompext
[X2go-Dev] 3.99 Testing - stray deletes
Hello, all. At first I thought this was my imagination but it is happening regularly and is a serious problem. Since upgrading to 3.99 on Debian Squeeze, I am getting lots of random delete key activations. I'll be reading a mail and it will suddenly disappear or a couple of mails will disappear from the Inbox. I then unhide deleted mails and see they've been deleted. I'll have some code open or a document and I'll suddenly see the document changed indicator activated. I'll look and there are missing letters. It's not even rare - it is a many times a day event. Has anyone else noticed this? - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Wed, 2011-07-27 at 09:33 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 14:41 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 14:33, schrieb Mike Gabriel: Hi Alex, On Mi 27 Jul 2011 13:52:46 CEST Oleksandr Shneyder wrote: Am 22.07.2011 06:38, schrieb John A. Sullivan III: I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hello John, Unfortunately, I can't reproduce this. I was able to start a several sessions simultaneously on debian and win7, see screen shots: http://www.x2go.org/screenshots/multisess-debian.png http://www.x2go.org/screenshots/multisess-win7.png Thanks for all your testing If I recall John correctly, her meant: start x2goclient, connect to session A as user A start another x2goclient, also connect to session A as user A (that is a second session as the same user to the same server) John, could you please cross-check this? Thanks, Mike As I understood, John meant, connecting with user A to server A, start second client and connecting as user B to server B. But it does not matter, I've tried all variants: A-A B-B A-A B-A A-A A-B A-A A-A regards, alex snip Oleksandr is correct - two separate users connected to two separate X2Go servers. I can do the initial connection but, within a few seconds, the first session times out. I've seen this myself on Debian and Windows XP and another tester confirmed it on Windows 7 64 bit. I'll try to dig out more details. Thanks very much for investigating - John snip OK - now this is bizarre - after days of it not working on Squeeze, it decided to work today when I tested. All right, who changed things while I wasn't looking! I'll give Windows a test later but I do know I wasn't imagining this and we had another tester confirm the same phenomenon. I'll check again with them, too. Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - Multiple X2goclient instances fail
On Wed, 2011-07-27 at 12:18 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 09:33 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 14:41 +0200, Oleksandr Shneyder wrote: Am 27.07.2011 14:33, schrieb Mike Gabriel: Hi Alex, On Mi 27 Jul 2011 13:52:46 CEST Oleksandr Shneyder wrote: Am 22.07.2011 06:38, schrieb John A. Sullivan III: I believe I have tested this on both Linux and Windows - becoming a bit of a blur after a long hot day!. In previous versions of X2GoClient, I could open multiple instances and connect to multiple different X2Go Server simultaneously. It was not official and it generated a few errors but it worked. This is actually a highly desirable feature for a possible very large account of ours. For some reason, that has stopped working in 3.99. It did not work in 3.0.1.18 either. It does work in Pyhoca so I do not think it is an impossibility. If at all possible, we would like to see this functionality restored - preferably officially but even accidentally as it was would do :) Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev Hello John, Unfortunately, I can't reproduce this. I was able to start a several sessions simultaneously on debian and win7, see screen shots: http://www.x2go.org/screenshots/multisess-debian.png http://www.x2go.org/screenshots/multisess-win7.png Thanks for all your testing If I recall John correctly, her meant: start x2goclient, connect to session A as user A start another x2goclient, also connect to session A as user A (that is a second session as the same user to the same server) John, could you please cross-check this? Thanks, Mike As I understood, John meant, connecting with user A to server A, start second client and connecting as user B to server B. But it does not matter, I've tried all variants: A-A B-B A-A B-A A-A A-B A-A A-A regards, alex snip Oleksandr is correct - two separate users connected to two separate X2Go servers. I can do the initial connection but, within a few seconds, the first session times out. I've seen this myself on Debian and Windows XP and another tester confirmed it on Windows 7 64 bit. I'll try to dig out more details. Thanks very much for investigating - John snip OK - now this is bizarre - after days of it not working on Squeeze, it decided to work today when I tested. All right, who changed things while I wasn't looking! I'll give Windows a test later but I do know I wasn't imagining this and we had another tester confirm the same phenomenon. I'll check again with them, too. Thanks - John snip The bizarreness continues. Now it works in Windows, too. Still waiting to hear back from the other tester. Other things have suddenly started to work, too, as I'll report separately, yet we've done nothing to change anything in our environment :( ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Wed, 2011-07-27 at 10:44 -0400, John A. Sullivan III wrote: On Wed, 2011-07-27 at 13:47 +0200, Oleksandr Shneyder wrote: snip Thanks, Alex, for taking the time to investigate. I'll comment in the text - John Hello all, last days I spent some time to test the latest version of x2goclient (3.99.0) on different platforms: Server debian lenny i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x debian squeeze amd64, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x Client debian lenny i386 debian squeeze i386 debian squeeze amd64 windows7-32 windows7-64 I've tested connections in LAN and over Internet and was able to connect and work without any serious issues. snip Unfortunately, I can not reproduce some bugs reported of users on this mailing list - Issues with image packaging, Issues with scrolling in firefox In my reporting, these two issues are related. The slow Firefox scrolling is from using pack=none. As I mentioned, the packing problem may not be strictly 3.99 but it is definitely a problem. I have it on Windows XP 32 bit, Windows 7 64 bit, Windows Server 2008 64 bit. I can confirm all three connecting to 3.0.1-5 X2GoServer and Windows XP on both 3.0.1-5 and 3.99.0.0. The XP system was a complete rebuilt from the recovery CD with all the latest patches applied (for hours and hours :( ). It did appear to work correctly with 3.0.1-5 server until I installed PyHoca. That's when the problems started. However, the Windows 7 and Windows Server 2008 systems never had Pyhoca installed. They have a fresh installation of 3.99.0. Oops, 2008 had 3.0.1-4. I just removed it, deleted all related files from Program Files, cleaned out the registry of every reference to x2go and obviouslynice and installed 3.99.0. Same problem immediately. Issues with delays after pressing Up/Down arrow keys. This is not so much an issue with Up/Down arrow keys. I was using it as an example of how dramatically slower X2Go is on Windows. I was using email selection in Evolution on a large Inbox to illustrate the point as Evolution has particularly bad screen handling. snip I can imagine, that some of this bugs are system dependent. X2Go uses many components - x2go server, client, nxlibs, nx(x2go)agent. Under windows x2goclient starts although X-server, SSH server and Pulseaudio server. All this components are depended from many system libraries, for example libpng, libjpeg, etc. Unfortunately, we can not to test X2Go with all possible combinations of software installed on user computer. We running our tests on fresh installed OS with default set of software. And default options of x2goclient (on Windows with default X-server and default command line options) If you expecting troubles running X2Go, please try, if it possible, to install a clean OS, without any additional software (f.e. third-part firewalls or antivirals on Windows) If X2Go running on such clean system, you can compare it with you productive system and find out which software blocking functionality of X2Go. As example, one of our customers found out, that x2go client on Win7 can not connect to server if Avast antivirus is installed. Quite understood and, as I think you know, I'll do whatever I can do to help but we do need to solve these issues. The packing one is critical. It may be a particular combination of packages and I'll enumerate those next as best I can. However, we can tell our clients to reinstall the client but we can't very well tell them to reinstall their OS to ensure it is clean. We'll need to find out what combination causes this so we can both correct it and warn against it. I've already mentioned the client sides and the Windows packaging is pretty self contained. Here's what I see on the server side: 3.0.1-5: ii x2goagent 3.4.0-3-1 Special NX-ized client, was developed by NoMachine to handle X sessions un x2goclientnone (no description available) ii x2gognomebindings 2.0.1-1 Description: x2gognomebindings (GNOME mimetypes for x2go) ii x2gokdebindings 2.0.1-4 x2gokdebindings (KDE mimetypes and icons for x2go) ii x2goprint 3.0.0-1 x2goprint command ii x2goserver3.0.1-5
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Wed, 2011-07-27 at 13:47 +0200, Oleksandr Shneyder wrote: Am 22.07.2011 12:10, schrieb Mike Gabriel: Hi all, thanks to everyone (esp. to John) for taking such a meticulous look at the latest x2goclient 3.99.0.0 (Win32 version). There obviously seem to be quite some open issues that we have to cross-check on our own development/testing systems. Alex (the core developer of x2goclient) will do that ASAP, we hope to come up with some feedback, patches, comments by the end of next week. If that gets delayed, I will let you know about the delay and come up with another timeline. Unfortunately, as X2go is not a commercial project, this cannot be any quicker. We apologize for this!!! If there are people out there you can read/write C code, please feel free to clone git://code.x2go.org/x2goclient.git and cross-read the code. Some other problems, however, may arise from x2goagent, as well. So you might use git://code.x2go.org/x2goagent.git for lecture as well... Greets, Mike Hello all, last days I spent some time to test the latest version of x2goclient (3.99.0) on different platforms: Server debian lenny i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze i386, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x debian squeeze amd64, x2goserver-3.0.99-5, nxlibs-3.4.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.4.x debian squeeze amd64, x2goserver-3.99.0.0, nxlibs-3.5.x, x2goagent-3.5.x Client debian lenny i386 debian squeeze i386 debian squeeze amd64 windows7-32 windows7-64 I've tested connections in LAN and over Internet and was able to connect and work without any serious issues. During tests I've found some bugs: - x2goclient crashing some times if user login unsuccessful (f.e. wrong password) - x2goclient crashing by unexpected output in user console (f.e. there is a line in ~/.basrc cowsay hello or user shell set to /bin/false ) - under Windows system tray icon does not disappear on client close. I'll fix this bugs ASAP. Unfortunately, I can not reproduce some bugs reported of users on this mailing list - Issues with image packaging, Issues with scrolling in firefox Issues with delays after pressing Up/Down arrow keys. Issues with failed connection on Win7-64 I can imagine, that some of this bugs are system dependent. X2Go uses many components - x2go server, client, nxlibs, nx(x2go)agent. Under windows x2goclient starts although X-server, SSH server and Pulseaudio server. All this components are depended from many system libraries, for example libpng, libjpeg, etc. Unfortunately, we can not to test X2Go with all possible combinations of software installed on user computer. We running our tests on fresh installed OS with default set of software. And default options of x2goclient (on Windows with default X-server and default command line options) If you expecting troubles running X2Go, please try, if it possible, to install a clean OS, without any additional software (f.e. third-part firewalls or antivirals on Windows) If X2Go running on such clean system, you can compare it with you productive system and find out which software blocking functionality of X2Go. As example, one of our customers found out, that x2go client on Win7 can not connect to server if Avast antivirus is installed. Hi, Alex. Although the packing and multi-session problems have mysteriously vaporized as I reported elsewhere, I can confirm that we still have the following: 1) Full screen toggles do not work in Windows - no magic pixel, AltCtl-M, AltCtl-F. 2) AltTab does not work as expected in Full Screen mode - in fact, rather than scroll through the X2Go desktop applications, it is currently the only way to break out of full screen! 3) I get a knotify: no process found error when exiting the Windows session. 4) Evolution selection (as a representative graphics issue) is still painfully slow Thanks - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] core dev-team testing x2goclient during next week
On Wed, 2011-07-27 at 06:52 -0700, John Williams wrote: On Wed, Jul 27, 2011 at 4:47 AM, Oleksandr Shneyder oleksandr.shney...@obviously-nice.de wrote: Unfortunately, we can not to test X2Go with all possible combinations of software installed on user computer. We running our tests on fresh installed OS with default set of software. And default options of x2goclient (on Windows with default X-server and default command line options) That is why there needs to be some way to turn on detailed logging so that people testing on other configurations can send logs to the developers to help find bugs. In my opinion, detailed logging should be a high priority for the clients, especially the Windows client. You did not mention the freezing bug that me and the other John saw. It seems like it may be related to AJAX. Any idea about what may be causing that one and how to fix it? snip I can confirm this is still a very serious problem. I was still in Windows when someone started chatting with me on Zimbra on my X2Go desktop. It literally took me 16 -2 20 second before I could respond to them after every single time they typed something even if the chat window was hidden. Needless to say, it made it excruciating to conduct that conversation. It then dawned on me to connect to Zimbra directly via the Internet bypassing X2Go. I was able to carry on the same conversation with no problem whatsoever - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] [PATCH] Allow users to edit their *own* sessions only
On Mon, 2011-07-25 at 08:32 +0200, Mike Gabriel wrote: Hi Reinhard, On Mo 25 Jul 2011 00:10:03 CEST Reinhard Tartler wrote: previously, users could create sessions under wrong uids or delete sessions from other users. This patch implements prevents this by checking the userid of the caller with the session id. +1 from me... [... patch ...] snip We addressed this a little differently as it is one of the problems we immediately recognized in X2Go two years ago and one of the major modifications we made in our environment. I'll have to dig out the specifics and your solution may be much better anyway but to scale to a large installation with a single database server and do it securely and without the users using the superuser database account, we changed all the scripts to use schemas named after the user's id. Each user has a schema and within the schema there is a trigger to update an instance of x2gosessions which is accessible by postgres. This table is used by a single x2gocleansessions routine which cleans up after all users rather than having 1000 such session all running every five seconds. The end result is a single database and a single cleanup daemon for an unlimited number of x2go servers and users with users having access to only their schema and no user using the postgres account - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] [PATCH] Allow users to edit their *own* sessions only
On Mon, 2011-07-25 at 13:19 +0200, Mike Gabriel wrote: Hi John, On Mo 25 Jul 2011 12:32:44 CEST John A. Sullivan III wrote: On Mon, 2011-07-25 at 08:32 +0200, Mike Gabriel wrote: Hi Reinhard, On Mo 25 Jul 2011 00:10:03 CEST Reinhard Tartler wrote: previously, users could create sessions under wrong uids or delete sessions from other users. This patch implements prevents this by checking the userid of the caller with the session id. +1 from me... [... patch ...] snip We addressed this a little differently as it is one of the problems we immediately recognized in X2Go two years ago and one of the major modifications we made in our environment. The script Reinhard modified only concerns SQLite... However, based on the current x2goserver code... if you come up with postgres patches (I strongly guess that you were refering to postgres changes on the x2goserver script) we will be happy to introspect them. Unfortunately, there has been a complete rewrite of the database scripts of X2go in the meantime, so I suppose your patches that you sent to the ML 1-2 years ago will not apply anymore. snip Yes, it is actually a high priority project for me to adapt to the new scripts and issue our new Squeeze desktops to replace our Lenny ones. It's amazing how many even higher priority projects keep intruding! - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
[X2go-Dev] 3.99 Testing - Return from full screen mode not working
This is also a Windows only issue. When in full screen mode there is no way to exit it other than AltTab. The standard ways of leaving full screen (magic pixel in top right, Alt-Ctl-F, or Alt-Ctl-M) do not work. Hmm . . . come to think of it, I do not think I tried Alt-Ctl-Enter but I am not in Windows now to test it - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev
Re: [X2go-Dev] 3.99 Testing - General sluggishness
On Fri, 2011-07-22 at 00:26 -0400, John A. Sullivan III wrote: I suppose I'll cover this one next as John Williams raised one of the issues. This is another Windows only issue. This issue has two categories. It is complicated by the packing problem because we can only run the Windows client now with sub-optimal packing but it was a problem before the corruption issue struck. Screen painting is just slower - not just a little slower - a lot slower. The typical anecdotal test I've been using is scrolling through my Evolution Inbox. Evolution has always seemed to have terrible graphics handling. I don't know what it does but it is always sluggish and awkward - thus an excellent test of how well we handle challenging graphics. Even under Linux on a LAN, scrolling is miserably slow. Shift down arrow and wait 1/4 second for a response. This is really annoying. However, in Windows, it is wait 1/2 to 3/4 seconds for a response. This takes it from annoying to dangerous as users wind up deleting or opening unintended emails, i.e., the hold the shift down arrow until they think they have highlighted the mails they want and then press enter. What they don't realize is that five more emails will be selected because of the delay between what they see and the key buffer. snip Ugh! Make that 2 to 3 seconds between down arrow and response I tried to resume a session I had suspended from Linux but it complained about a mismatched color depth. I checked my Windows laptop and, indeed, it was set to 16 bit color. I changed it to 32 bit color and performance has gone through the floor - John ___ X2go-Dev mailing list X2go-Dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/x2go-dev