The System.exit() issue should be fixed in trunk, as well as the other issues related to changing parameters in the New Connection dialog, etc. New build is forthcoming once I iron out some issues with full-screen mode. The delay is still a mystery, but it occurs with TigerVNC as well (completely different web server implementation.) The fact that it doesn't occur with any other plugins leads me to believe it's an IceaTea bug. I'm perfectly willing to change something in our JAR if it would work around it, but I have no idea what.
On 8/9/12 2:48 AM, DRC wrote: > I can reproduce this readily on RHEL 6. The delay is due to an unknown > issue in the plugin, and a Google search reveals that we don't seem to > be the only ones experiencing that. The other issue, however, is a bug > in our code-- we shouldn't be calling System.exit() in applet mode. > Investigating that. > > > On 8/8/12 9:55 PM, Shanon Loughton wrote: >> For my a) b) and c) I tried removing everything in cache in >> ~/.icedtea/cache and also Icedtea Web Control Panel, restarted >> everything and still the same. It was definitely the new interface as >> the Server popup box is the new style. The stdout/stderr looks like >> this: >> >> (Type address bar http://localhost:5801 [enter ]) >> >> java version "1.7.0_03" >> OpenJDK Runtime Environment (IcedTea7 2.1.1pre) >> (7~u3-2.1.1~pre1-1ubuntu3) >> OpenJDK 64-Bit Server VM (build 22.0-b10, mixed mode) >> >> (2 minute wait in here..., then->) >> >> CConn: connected to host localhost port 5901 >> CConnection: Server supports RFB protocol version 3.8 >> CConnection: Using RFB protocol version 3.8 >> >> (popup Server, change anything, click 'OK' ) >> >> Exception in thread "Thread-13" java.security.AccessControlException: >> Applets may not call System.exit() >> at >> net.sourceforge.jnlp.runtime.JNLPSecurityManager.checkExit(JNLPSecurityManager.java:385) >> >> at java.lang.Runtime.exit(Runtime.java:105) >> at java.lang.System.exit(System.java:960) >> at com.turbovnc.vncviewer.CConn.deleteWindow(CConn.java:202) >> at com.turbovnc.vncviewer.VncViewer.run(VncViewer.java:281) >> at java.lang.Thread.run(Thread.java:722) >> >> And java plugin process is still running too, no popup or window anymore. >> >> The issue is definitely between OpenJDK and Icedtea plugin. I get the >> same issue on openjdk 1.6: >> java version "1.6.0_24" >> OpenJDK Runtime Environment (IcedTea6 1.11.3) >> (6b24-1.11.3-1ubuntu0.12.04.1) >> OpenJDK 64-Bit Server VM (build 20.0-b12, mixed mode) >> >> These symptoms do not exist with Oracle Java version >> java version "1.7.0_04" >> Java(TM) SE Runtime Environment (build 1.7.0_04-b20) >> Java HotSpot(TM) 64-Bit Server VM (build 23.0-b21, mixed mode) >> and ensuring the Oracle Java plugin /usr/lib/mozilla/plugins/libnpjp2.so >> -> /usr/lib/jvm/jdk1.7.0_04/jre/lib/amd64/libnpjp2.so >> I installed the above from here : >> http://www.oracle.com/technetwork/java/javase/downloads/jdk-7u4-downloads-1591156.html >> >> >> As for d) you're right its a default and pre-existing behaviour. I >> guess just need play with the html wrap code to understand it better. >> >> cheers >> Shanon >> >> >> On 8 August 2012 15:59, DRC <dcomman...@users.sourceforge.net >> <mailto:dcomman...@users.sourceforge.net>> wrote: >> >> On 8/6/12 2:43 AM, Shanon Loughton wrote: >> > a) From the browser address bar, the vncviewer can take a while >> to load, >> > up 2 minutes. See the 3 lines' timestamp in the server vnc log >> "httpd: >> > get 'VncViewer.jar'". Even locally takes this long, whereas >> before was >> > fairly instantaneous. >> > >> > 06/08/2012 12:26:01 httpd: get '' for xxx.xx.xx.xxx >> > 06/08/2012 12:26:01 httpd: defaulting to 'index.vnc' >> > 06/08/2012 12:26:07 httpd: get 'VncViewer.jar' for xxx.xx.xx.xxx >> > 06/08/2012 12:28:08 httpd: get 'VncViewer.jar' for xxx.xx.xx.xxx >> > 06/08/2012 12:28:08 httpd: get 'VncViewer.jar' for xxx.xx.xx.xxx >> > >> > 06/08/2012 12:28:16 Got connection from client xxx.xx.xx.xxx >> > 06/08/2012 12:28:16 (other clients xxx.xx.xx.xxx) >> > 06/08/2012 12:28:16 Using protocol version 3.8 >> > 06/08/2012 12:28:16 Enabling TightVNC protocol extensions >> >> I haven't seen this, but I have seen a variety of other issues with >> loading VncViewer.jar through the built-in HTTP server. I finally >> tracked them all down, and the cause was that, for whatever >> reason, the >> Java plug-in had cached an old version of the TigerVNC Java Viewer >> (which is also called VncViewer.jar.) When it tried to load that >> viewer >> from the cache, it would briefly display a dialog asking me to >> validate >> the TigerVNC self-signed key, which confused me to no end, since I >> couldn't figure out where it was picking up that key. >> >> In the long term, it might be prudent for me to rename our viewer to >> avoid such issues in the future, but at least for now, please try >> clearing our your JAR cache (this is in the Java preferences. On my >> machines, at least, it's in General->Settings on Windows and under >> the >> "Network" tab on Mac.) >> >> >> > b) Once loaded, I get the java vncviewer popup with Server field >> and >> > buttons for 'options', 'ok' etc.. The Server field shows the >> server IP >> > address and ":1". Shouldnt it be :5801? Neither of them work, >> once you >> > press OK it closes/crashes silently. >> > Where can I check the client log? >> >> No, :1 is correct. 5801 is the HTTP port, so that's what you >> enter into >> your browser. The VNC connection uses port 5901, however, which >> corresponds to display :1. >> >> >> > c) If I change anything in the Server field, then click Options, >> it also >> > closes/crashes silently. >> >> Can't reproduce that. Try clearing your JAR cache. >> >> >> > d) Firefox is not prompted to install a java applet plugin when >> browsing >> > tohttp://ipaddress:5801, instead a link to the Turbovnc site is >> only >> shown. >> > Is this because when called from a browser, it should be called >> from an >> > embedded object, rather than directly through the address bar? >> > Workaround: install icedtea 6 or 7 firefox plugin through apt-get. >> >> Has that behavior changed since 1.1? I suspect not. index.vnc is >> loading the viewer through an <APPLET> tag. If there is another way >> that it can be called that would force the plug-in to be installed >> (possibly an <EMBED> tag), I'm willing to entertain the notion of >> changing index.vnc, but I don't have time to fool with that >> myself. If >> this doesn't represent a regression, then it's not something I can >> fix >> right now. >> >> >> > Moreover when called from the address bar, the server vnc log >> shows: >> > 06/08/2012 12:22:11 httpd: get '' for yyy.yy.yy.yyy >> > 06/08/2012 12:22:11 httpd: defaulting to 'index.vnc' >> > 06/08/2012 12:22:11 httpd: get >> 'com.turbovnc.vncviewer.VncViewer' for >> > yyy.yy.yy.yyy >> > 06/08/2012 12:22:11 httpProcessInput: open: No such file or >> directory >> > >> > which is probably incorrect. >> >> Not sure why that would happen unless the browser is sending an >> incorrect GET request. I haven't seen it, and I'm using Firefox (but >> not the same Java plug-in as you are.) I am really strapped for >> time to >> debug this, so any help that the community can give is greatly >> appreciated. >> >> >> ------------------------------------------------------------------------------ >> >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. >> Discussions >> will include endpoint security, mobile security and the latest in >> malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> VirtualGL-Users mailing list >> VirtualGL-Users@lists.sourceforge.net >> <mailto:VirtualGL-Users@lists.sourceforge.net> >> https://lists.sourceforge.net/lists/listinfo/virtualgl-users >> >> >> >> >> ------------------------------------------------------------------------------ >> >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> >> >> >> _______________________________________________ >> VirtualGL-Users mailing list >> VirtualGL-Users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/virtualgl-users >> ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ VirtualGL-Users mailing list VirtualGL-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtualgl-users
