Azureus hangs or crashes showing splash screen at start

Confirmed, I think. There's more than one problem to fix before azureus will run.

Firstly, in /usr/bin/azureus, two of the JAR file references are wrong - there should be references to glib0.4.jar and gtk2.10.jar in $CP.

Secondly, once that is changed, I get the backtrace
  java.lang.UnsatisfiedLinkError: gtk_icon_theme_get_default
        at org.gnu.gtk.IconTheme.gtk_icon_theme_get_default(Native Method)
        at org.gnu.gtk.IconTheme.getDefault(IconTheme.java:42)
        at org.gudy.azureus2.ui.swt.ImageRepository.getThemedIcon(ImageRepository.java:71)
        at org.gudy.azureus2.ui.swt.ImageRepository.loadImages(ImageRepository.java:128)
        at org.gudy.azureus2.ui.swt.mainwindow.Initializer.run(Initializer.java:210)
        at org.gudy.azureus2.ui.swt.mainwindow.SWTThread$1.runSupport(SWTThread.java:107)
        at org.gudy.azureus2.core3.util.AERunnable.run(AERunnable.java:38)
        at java.lang.Thread.run(Unknown Source)

Not sure what's happening here - /usr/lib/jni/libgtk-2.10.so has a DT_NEEDED entry for libgtk-x11-2.0.so.0 - so I think that SHOULD cause the GTK library to be linked in? Unsure if this is a problem with libgtk-java, or...

note: using SUN Java 1.5.0+update05

If using gij/gcj, azureus WILL start if LD_PRELOAD is set to /usr/lib/jni/libgtkjni-2.10.so:/usr/lib/jni/libglibjni-0.4.so

With gcj-4.1 (4.1.1-11ubuntu7), azureus works without any LD_PRELOAD set (but with the other changes to /usr/bin/azureus mentioned above).

I also see different behaviour with SUN java:
  java.lang.UnsatisfiedLinkError: /usr/lib/jni/libglibjni-0.4.so: /usr/lib/jni/libglibjni-0.4.so: undefined symbol: g_type_depth
        at java.lang.ClassLoader$NativeLibrary.load(Native Method)
        at java.lang.ClassLoader.loadLibrary0(Unknown Source)
        at java.lang.ClassLoader.loadLibrary(Unknown Source)
        at java.lang.Runtime.loadLibrary0(Unknown Source)
        at java.lang.System.loadLibrary(Unknown Source)
        at org.gnu.glib.Struct.<clinit>(Struct.java:85)
        at org.gudy.azureus2.ui.swt.ImageRepository.getThemedIcon(ImageRepository.java:71)
        at org.gudy.azureus2.ui.swt.ImageRepository.loadImages(ImageRepository.java:128)
        at org.gudy.azureus2.ui.swt.mainwindow.Initializer.run(Initializer.java:210)
        at org.gudy.azureus2.ui.swt.mainwindow.SWTThread$1.runSupport(SWTThread.java:107)
        at org.gudy.azureus2.core3.util.AERunnable.run(AERunnable.java:38)
        at java.lang.Thread.run(Unknown Source)

The DT_NEEDED information from libglibjni is:
$ objdump -x /usr/lib/jni/libglibjni-0.4.so | grep NEEDED
  NEEDED libglib-2.0.so.0
  NEEDED libc.so.6

But g_type_depth is not defined there... setting LD_PRELOAD to /usr/lib/libgobject-2.0.so works in this case. Is there an --as-needed in a linking step somewhere?

Another confirmation, using sun java 1.5.0-08-0ubuntu1:

rikai@sonic:~$ azureus
changeLocale: *Default Language* != English (United States). Searching without country..
changeLocale: Searching for language English in *any* country..
changeLocale: no message properties for Locale 'English (United States)' (en_US), using 'English (default)'
DEBUG::Fri Sep 01 18:55:00 EDT 2006::org.gudy.azureus2.ui.swt.mainwindow.Initializer::run::304:
  java.lang.NoClassDefFoundError: org/gnu/gtk/IconTheme
        at org.gudy.azureus2.ui.swt.ImageRepository.getThemedIcon(ImageRepository.java:71)
        at org.gudy.azureus2.ui.swt.ImageRepository.loadImages(ImageRepository.java:128)
        at org.gudy.azureus2.ui.swt.mainwindow.Initializer.run(Initializer.java:210)
        at org.gudy.azureus2.ui.swt.mainwindow.SWTThread$1.runSupport(SWTThread.java:107)
        at org.gudy.azureus2.core3.util.AERunnable.run(AERunnable.java:38)
        at java.lang.Thread.run(Thread.java:595)

+1 here too. Broken with the same exception.

+1 here...

Same here. Changes to /usr/bin/azureus and setting LD_PRELOAD according to the first solution here got me up.

Updating now, will see how it goes.

Yeah well... It does not start with Sun Java...
With gcj, on the other hand, transfers stop after less than an hour.

on the other hand you can download azureus from sourceforge and run the jar file. It works fine with sun java. Personally dont use the other javas...

on the other hand you can download azureus from sourceforge and run the shell script. It works fine with sun-java5-jdk . Personally I dont use the other javas and removed them...

+1

All the same symptoms.

Same problem o edgy amd64.
With gcj downloads stop after a while.

Azureus also doesn't load for me, using gcj on edgy amd64

I'm on Ubuntu Edgy - Latest ( Beta ).
The splash screen show *loading images* and keeps displaying that until I force-quit it.
Here's what my console displays:

kristof@Edgy:~$ azureus
DEBUG::Sat Oct 07 15:21:41 GMT+02:00 2006::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::137:
  No SSL provider available
    SESecurityManager::initialise::52,ConfigurationChecker::setSystemProperties::141,ConfigurationManager::initialise::138,ConfigurationManager::getInstance::71,LoggerImpl::init::90,Logger::<clinit>::48,Class::initializeClass::-1,StartServer::<init>::69,Main::<init>::56,Main::main::162
changeLocale: *Default Language* != en (AU). Searching without country..
changeLocale: Searching for language en in *any* country..
changeLocale: no message properties for Locale 'en (AU)' (en_AU), using 'English (default)'
DEBUG::Sat Oct 07 15:21:42 GMT+02:00 2006::org.gudy.azureus2.ui.swt.mainwindow.Initializer::run::304:
  java.lang.NoClassDefFoundError: org.gnu.gtk.IconTheme
   at org.gudy.azureus2.ui.swt.ImageRepository.getThemedIcon(ImageRepository.java:71)
   at org.gudy.azureus2.ui.swt.ImageRepository.loadImages(ImageRepository.java:128)
   at org.gudy.azureus2.ui.swt.mainwindow.Initializer.run(Initializer.java:210)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread$1.runSupport(SWTThread.java:107)
   at org.gudy.azureus2.core3.util.AERunnable.run(AERunnable.java:38)
   at java.lang.Thread.run(libgcj.so.70)

I'm getting the same error on AMD64 Edgy Beta.

I've also tried Azureus 2.5 downloaded from Sourceforge rather than the 2.4 version in the repositories, and I got the same error.

Updated list of things required to get the azureus package to work:
1. In debian/azureus.sh, change "glib0.2.jar" to "glib0.4.jar".
2. In debian/azureus.sh, change "gtk2.8.jar" to "gtk2.10.jar".
3. In debian/azureus.sh, add "export LD_LIBRARY_PATH=/usr/lib/jni:/usr/lib".
4. Add a call to org.gtk.Gtk.init before org.gudy.azureus2.ui.swt.ImageRepository.loadImages is called. The problem here is that the GTK library has not yet been loaded, so none of the GTK calls can be resolved. In my testing, I edited azureus/org/gudy/azureus2/ui/swt/mainwindow/Initializer.java, and added the following lines to the run() method:
        String gtk_args[] = new String[1];
        gtk_args[0] = "azureus";
        org.gnu.gtk.Gtk.init(gtk_args);

The call can actually go anywhere as long as it happens before any GTK calls are made.

Thanks Gary - that's working for me.

I got the 2.5.0.0 source from Azureus CVS, made the change to Initializer.java and built with Ant. Then copied Azureus2.jar to /usr/share/java/ (where the apt installed one lives).

Made the changes you suggested to /usr/bin/azureus, ran it and it works.

I have this same problem on Kubuntu 6.10. I solved the problem by replacing /usr/share/java/Azureus.jar with Azureus2.5.0.0.jar (rename to Azureus.jar) which can be downloaded from http://prdownloads.sourceforge.net/azureus/Azureus2.5.0.0.jar?download

When Azureus 2.5.0.0 is started the first time, it gives a warning notice about old Eclipse but it does not come up afterwards and as far as I can tell, works just fine.

I would expect the upstream version to work.

The core of the problem with the Ubuntu package is that one of the branding patches (it tries make certain UI icons fit the current GTK+ theme) makes calls into GTK before the Java GTK libraries have been fully initialized. It does not appear to be in use upstream.

An alternative to the above change to Initializer.java is to remove the patch "azureus-themed.patch" from the Ubuntu package.

Confirming that azureus does not work properly when installed from the ubuntu repo. I installed 2.5.0.0 from the source, and put it in /usr/share/azureus and it works fine as long as a I run it with the full path. The script in /usr/bin needs to be replaced, and azureus should be upgraded from 2.4 to 2.5. Anyway, thats how I worked around this issue, hopefully we'll see it fixed before Edgy is deemed stable.

You would think these types of bugs can never hit the repos. They have this patch. They apply it. They upload it. They never ever try to run it.

I'm not expecting any advanced regression test suite or anything. But at least try to run it once.

PS. Its now 23/10. Is the fix proposed above going to get in on time before the release?

The status and purpose of this package may be questionable, but the movie-downloading Azureus types are one of Ubuntu's strong backers. They're the ones setting linux up for friends and family and are their first line of support.

Getting a nice working Azureus by default is more important and noticeable to the 'general-public' than we might at first hand expect. Although I don't think we'll see any reviews complaining about this though ;-)

+1

fixed in azureus_2.5.0.0-0ubuntu2

Really fixed? I have updated to azureus_2.5.0.0-0ubuntu2. The problem is still occurring with sun hotspot 1.5.09. I've tried with azureus-gcj and all dependencies, but the error

Exception in thread "main" java.lang.UnsatisfiedLinkError: memmove
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:67)
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:54)
   at org.eclipse.swt.widgets.Display.<clinit>(Display.java:126)
   at java.lang.Class.initializeClass(libgcj.so.70)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.createInstance(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.Initializer.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.main(Azureus2.jar.so)

insists in occur.

yes, that's amd64 specific, see #68053

Nope, still does not start here.

variant@craptop:~$ azureus
changeLocale: *Default Language* != English (Ireland). Searching without country..
changeLocale: Searching for language English in *any* country..
changeLocale: no message properties for Locale 'English (Ireland)' (en_IE), using 'English (default)'
#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# SIGSEGV (0xb) at pc=0xb0561d02, pid=15371, tid=3084916400
#
# Java VM: Java HotSpot(TM) Client VM (1.5.0_08-b03 mixed mode, sharing)
# Problematic frame:
# C [libglibjni-0.4.so+0x8d02]
#
# An error report file with more information is saved as hs_err_pid15371.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

this is celeron cpu, happens with official sun jdk or gcj

I report that Azureus version 2.5.0.0-0ubuntu2 works perfectly in my 32-bit Edgy (upgraded from Dapper).

Azureus crash on Edgy (AMD Athlon):

user:~/download$ java -version
java version "1.5.0_09"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_09-b03)
Java HotSpot(TM) Client VM (build 1.5.0_09-b03, mixed mode, sharing)
user:~/download$ azureus
changeLocale: *Default Language* != English (Canada). Searching without country..
changeLocale: Searching for language English in *any* country..
changeLocale: no message properties for Locale 'English (Canada)' (en_CA), using 'English (default)'
#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# SIGSEGV (0xb) at pc=0xb096ed02, pid=17232, tid=3085346480
#
# Java VM: Java HotSpot(TM) Client VM (1.5.0_09-b03 mixed mode, sharing)
# Problematic frame:
# C [libglibjni-0.4.so+0x8d02]
#
# An error report file with more information is saved as hs_err_pid17232.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

Kien Tran, I get this error too. Additionally I experience it on both on my machines at home (both running dapper).
The only way I found to temporarily resolve it is to rm -r the .azureus directory in my /home/user directory.

Erik, thank you for the tip. I did rm -r on .azureus in my home directory. and azureus started, and worked for a while, then POOF, it SEGV'ed again, at the same place. Seems like the culprit is libglibjni-0.4.so, but I don't know how to fix it. Here are the beginning of 2 crash logs. In retrospect, upgrading from Dapper to Edgy has brought me nothing but headaches.
First log:
#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# SIGSEGV (0xb) at pc=0xb0945d02, pid=2845, tid=3084777136
#
# Java VM: Java HotSpot(TM) Client VM (1.5.0_09-b03 mixed mode, sharing)
# Problematic frame:
# C [libglibjni-0.4.so+0x8d02]
#
Second log:
#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# Internal Error (53484152454432554E54494D450E43505001A3), pid=4814, tid=3085354672
#
# Java VM: Java HotSpot(TM) Client VM (1.5.0_09-b03 mixed mode, sharing)

--------------- T H R E A D ---------------

Current thread (0x0805cc50): JavaThread "main" [_thread_in_Java, id=4814]

Stack: [0xbf742000,0xbf942000), sp=0xbf93bcc0, free space=2023k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [libjvm.so+0x3251c3]
V [libjvm.so+0x12442c]
V [libjvm.so+0x2c257e]
V [libjvm.so+0x28ce24]
V [libjvm.so+0x28a034]
C [+0x440] __kernel_rt_sigreturn+0x0
V [libjvm.so+0x18629a]
C [libglibjni-0.4.so+0x8d51]
C [libglib-2.0.so.0+0x33f87] g_logv+0x327
C [libglib-2.0.so.0+0x34159] g_log+0x29
C [libgtk-x11-2.0.so.0+0x25590b] gtk_widget_size_allocate+0x2ab
C

Just a comment: I've installed Edgy and Azureus was working fine (even with a .azureus that I copied from a FC 3 machine) initially. But one I've installed Sun's JRE, it got broken with the aforementioned JVM crash.

From the stack trace, it's kind of obvious that the Java setup is messed up - the JVM being called is Sun's, but some GCJ libs are getting in the way. So, my first reaction was to uninstall Sun's or the GJC stuff, but ended up finding an easier solution: I've installed azureus-gcj, and everything works fine now..

Hello I have Edgy on amd64 clean install and I get this error

alo have installed azureus-gcj
________________________________

azureus
DEBUG::Wed Nov 08 18:11:00 GMT+01:00 2006::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
changeLocale: *Default Language* != en (US). Searching without country..
changeLocale: Searching for language en in *any* country..
changeLocale: no message properties for Locale 'en (US)' (en_US), using 'English default)'
DEBUG::Wed Nov 08 18:11:01 GMT+01:00 2006::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,CryptoManagerImpl::<init>::-1,CryptoManagerImpl::getSingleton::-1,CryptoManagerFactory::getSingleton::-1,AzureusCoreImpl::<init>::-1,AzureusCoreImpl::create::-1,AzureusCoreFactory::create::-1,Main::<init>::-1,Main::main::-1
Exception in thread "main" java.lang.UnsatisfiedLinkError: memmove
   at org.eclipse.swt.internal.gtk.OS.memmove(org.eclipse.swt.gtk.linux.x86_64_3.2.1.v3235.jar.so)
   at org.eclipse.swt.internal.Converter.wcsToMbcs(org.eclipse.swt.gtk.linux.x86_64_3.2.1.v3235.jar.so)
   at org.eclipse.swt.internal.Converter.wcsToMbcs(org.eclipse.swt.gtk.linux.x86_64_3.2.1.v3235.jar.so)
   at org.eclipse.swt.widgets.Display.<clinit>(org.eclipse.swt.gtk.linux.x86_64_3.2.1.v3235.jar.so)
   at java.lang.Class.initializeClass(libgcj.so.70)
   at org.eclipse.swt.widgets.Display.getCurrent(org.eclipse.swt.gtk.linux.x86_64_3.2.1.v3235.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.createInstance(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.Initializer.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.main(Azureus2.jar.so)

Azureus crashed with JRE 1.5 update 9 as well.

Installed azureus-gcj (x86).. works if I do "sudo update-alternatives --config java" and select java-gcj

Problem is, this breaks Frostwire...

So, altered /usr/bin/frostwire to be:

#!/bin/bash
cd /usr/lib/frostwire
export JAVA_PROGRAM_DIR=/usr/lib/jvm/java-1.5.0-sun/bin/
sh runFrost.sh

So for now - both work.... ah.. the joys of java

Sorry, but i have azureus_2.5.0.0-0ubuntu2 (edgy x86) and this bug is still present. After upgrade from Dapper i have problem described in posts. With gcj azureus working for a while and than stop. With sun java azureus crashing very often

Follow up to my previous posts:
Running with Sun Java 5: core dumps
Running with gjc: stops working after a few hours

Downloaded Azureus_2.5.0.0_linux.tar.bz2 from sourceforge.net, installed in home directory, ran it with Sun Java 5. No problems since.

Confirmed, still broken with Sun Java 5.

-It will run for a while, then crash
- Then next time you restart it, it won't even start anymore
- If you delete your ~/.azureus folder, it will run for a few hours, and crash again.
- Repeat.

This is really starting to become annoying, somebody should fix this completely and once and for all as soon as possible.

Console log from the crashes:

#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# Internal Error (53484152454432554E54494D450E43505001A3), pid=13462, tid=3084568240
#
# Java VM: Java HotSpot(TM) Client VM (1.5.0_08-b03 mixed mode, sharing)
# An error report file with more information is saved as hs_err_pid13462.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

reopening as it appears to still be an issue

DEBUG::Fri Jan 05 02:12:56 GMT+02:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
changeLocale: *Default Language* != en (ZA). Searching without country..
changeLocale: Searching for language en in *any* country..
changeLocale: no message properties for Locale 'en (ZA)' (en_ZA), using 'English (default)'
DEBUG::Fri Jan 05 02:12:57 GMT+02:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,CryptoManagerImpl::<init>::-1,CryptoManagerImpl::getSingleton::-1,CryptoManagerFactory::getSingleton::-1,AzureusCoreImpl::<init>::-1,AzureusCoreImpl::create::-1,AzureusCoreFactory::create::-1,Main::<init>::-1,Main::main::-1
Exception in thread "main" java.lang.UnsatisfiedLinkError: memmove
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:67)
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:54)
   at org.eclipse.swt.widgets.Display.<clinit>(Display.java:126)
   at java.lang.Class.initializeClass(libgcj.so.70)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.createInstance(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.Initializer.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.main(Azureus2.jar.so)

AMD64 3200+ Ubuntu edgy for AMD64, updated to latest packages. I tried the fixes above, did not work. java version "1.4.2"
gij (GNU libgcj) version 4.1.2 20060928 (prerelease) (Ubuntu 4.1.1-14ubuntu7)

Same bug, nearly same debug info:

DEBUG::Fri Jan 12 12:53:33 EST 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
changeLocale: *Default Language* != en (US). Searching without country..
changeLocale: Searching for language en in *any* country..
changeLocale: no message properties for Locale 'en (US)' (en_US), using 'English (default)'
DEBUG::Fri Jan 12 12:53:34 EST 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,CryptoManagerImpl::<init>::-1,CryptoManagerImpl::getSingleton::-1,CryptoManagerFactory::getSingleton::-1,AzureusCoreImpl::<init>::-1,AzureusCoreImpl::create::-1,AzureusCoreFactory::create::-1,Main::<init>::-1,Main::main::-1
Exception in thread "main" java.lang.UnsatisfiedLinkError: memmove
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:67)
   at org.eclipse.swt.internal.Converter.wcsToMbcs(Converter.java:54)
   at org.eclipse.swt.widgets.Display.<clinit>(Display.java:126)
   at java.lang.Class.initializeClass(libgcj.so.70)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.SWTThread.createInstance(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.mainwindow.Initializer.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.<init>(Azureus2.jar.so)
   at org.gudy.azureus2.ui.swt.Main.main(Azureus2.jar.so)

more info about the log problem can be found here https://launchpad.net/ubuntu/+source/azureus/+bug/68020

Am also waiting for this

Unbelieveable. After all this time this package is still broken. And the official upstream version just works for everybody.

You would expect at this point, that the packager would just give up, throw away his own patches, and use the upstream version that DOES WORK FOR EVERYBODY.

Ah, well, maybe the repository's version will start working somewhere in the next 5 years.

If the fix is as simple as 'stop using your own patches, just use upstream' then why isn't it fixed after all this time?

I don't want to sound mean: but at this point its just better to remove the azureus package from the universe because ubuntu is not able to offer such a package in a way that it actually works.

I have to agree. This bug should be considered a blocker for this package and it is better to remove it entirely from the repositories until it can be fixed.

It is unfortunate that a popular package such as this is unavailable due to what are presumably *downstream* issues.

The weird thing is, the testing before the initial release was zero. It didn't work with any JVM on any machine. But when the release is final, it becomes very difficult to patch a package. You will need to get all kinds of permissions and such.

I would say:
 - less bureaucracy when patching packages after final release.
 - more bureaucracy when adding a package to a release before its final

And if it doesn't work at release time, do not include the package at all. Maybe it'll be in time for the next ubuntu-release, but it missed this one.

Sort of something like: it requires 10 people to say 'yeah, it works' before it can even reach the final universe-repository. (before that it should just be in proposed like a patch is after the final release)

Considering the fact that its one of those programs at least 20% of all new users will install first (because they were also using it on windows), it makes a really bad impression of Ubuntu. Would it not be in the repositories, they would have downloaded it from the site, and it would have worked. However, will they know that the repository version is weird hacked up version? No, they won't. They won't try the upstream version, they will give up. Perhaps on a Ubuntu as a whole.

Yeah. Not being able to run a java package in a distribution is pretty lame. Presumably the dependencies are much more trivial than a c program, but still it crashes all the time in ubuntu.... shame really...

The new package in ubuntu-proposed launches for me (after "rm -r ~/.azureus") now. But there is still a problem with the 'welcome' screen and localization:

  Welcome Window: Error loading resource:
  /org/gudy/azureus2/internat/whatsnew/wat is er nieuw.Nederlands.stf

I haven't tried downloading with it though. I'm using Sun's Java at the moment.
Upstream does support dutch though, although it seems its not packaged here.

Maybe we should make it fall-back to english when the welcome-screen isn't available?

It might still be the case that the desktop is launched with a language not supported by Azureus localization. Is it managed by Rosetta?

I used automatix to install the java and azureus and now it works,
even if I try run BitTyrant which is based on azureus.

On 2/2/07, Ralf Nieuwenhuijsen <email address hidden> wrote:
> The new package in ubuntu-proposed launches for me (after "rm -r
> ~/.azureus") now. But there is still a problem with the 'welcome' screen
> and localization:
>
> Welcome Window: Error loading resource:
> /org/gudy/azureus2/internat/whatsnew/wat is er nieuw.Nederlands.stf
>
> I haven't tried downloading with it though. I'm using Sun's Java at the moment.
> Upstream does support dutch though, although it seems its not packaged here.
>
> Maybe we should make it fall-back to english when the welcome-screen
> isn't available?
>
> It might still be the case that the desktop is launched with a language
> not supported by Azureus localization. Is it managed by Rosetta?
>
> --
> Azureus does not start
> https://launchpad.net/bugs/57875
>

I have the same problem here with the current version of Ubuntu 6.10. I tried to install sun-java6-bin instead of Java 5 but then I get:

/usr/bin/azureus: line 36: exec: java: not found

Stephan, run

sudo update-alternatives --config java

and choose

/usr/lib/jvm/java-6-sun/jre/bin/java

I've found that deleting ~/.azureus/logs (rm -r ~/.azureus/logs) will allow me to open azureus again if it starts giving me the HotSpot error. This way I don't lose my settings and torrents. Running Edgy, sun-java6-jre on an Athlon XP.

Nice tip with deleting ~/.azureus-/logs. That worked fine for me too, and I was experiencing the same crashing problem as everyone else. Deleting the entire .azureus directory is annoying, since it loses the torrents. Hopefully the fact that deleting logs 'cures' this problem will be the clue that the ubuntu packager needs to get this problem resolved. Azureus is, IMO, the best bittorrent client available in the repositories, so would be nice if it worked well.

It works with Feisty (herd 4) and restricted-extra's package (which installs java)

Please, dear packaging-friends, keep it like this ;-)

It crashed for me too; I just deleted ~/.azureus, and now it works perfectly.

I don't think this bug is resolved, because that step should not be necessary; we can't expect everyone upgrading from Edgy who uses Azureus to do this.

Is it terribly hard to backport the fix to Edgy?

I'm still having this problem on Feisty.

It's easy to duplicate, just try to open a .torrent file using Azureus by right clicking -> Open with Azureus. It will crash and the after that Azureus will not start unless the ~/.azureus is deleted.

Using Java 6 from the repositories. Same story on three different laptops, and with Java 5.

apt-get install sun-java6-jre ia32-sun-java6-bin
update-alternatives --config java
 + 3 /usr/lib/jvm/java-gcj/jre/bin/java
* 4 /usr/lib/jvm/java-6-sun/jre/bin/java
worked for me

On 3/8/07, Thom Pischke <email address hidden> wrote:
> I'm still having this problem on Feisty.
>
> It's easy to duplicate, just try to open a .torrent file using Azureus
> by right clicking -> Open with Azureus. It will crash and the after
> that Azureus will not start unless the ~/.azureus is deleted.
>
> Using Java 6 from the repositories. Same story on three different
> laptops, and with Java 5.
>
> --
> Azureus does not start
> https://launchpad.net/bugs/57875
>

My solution of this problem is (I used Azureus from: Application -> Add/Remote): download from Azureus sourceforge page a new version for amd64 (my computer). It's work, when I unpaked and run.

(yes - english is not my native language)

Confirmed on Ubuntu Feisty (as of 3.4.2007) using azureus-2.5.0.0 and
Java 1.6.0
Sun Microsystems Inc.
SWT v3236, gtk
Linux v2.6.20-13-generic, i386

Removing the files in .azureus/logs makes it possible to start.

confirmed of Feisty.

even though removing the files in .azureus/logs makes it possible to start, it does crash soon after.

update-alternatives --config java
to see the current settings
apt-get install sun-java6-jre ia32-sun-java6
update-alternatives --config java
and change it to the new one

On 4/8/07, amitm02 <email address hidden> wrote:
> confirmed of Feisty.
>
> even though removing the files in .azureus/logs makes it possible to
> start, it does crash soon after.
>
> --
> Azureus does not start
> https://bugs.launchpad.net/bugs/57875
> You received this bug notification because you are a direct subscriber
> of the bug.
>

I'm already using Java6. It's not a 5 v. 6 problem

Fixed if for me, though I downloaded the Azureus binary from their
official website, I run it manually and using root. It is no solution,
but a temporary fix if you need it.

On 4/8/07, maco <email address hidden> wrote:
> I'm already using Java6. It's not a 5 v. 6 problem
>
> --
> Azureus does not start
> https://bugs.launchpad.net/bugs/57875
> You received this bug notification because you are a direct subscriber
> of the bug.
>

After updating to Feisty "rm -r ~/.azureus/logs" fixed the problem for me too, i.e. it starts up and I have no crashes since then. By the way is there still a chance of an upgrade to 2.5.0.4 for feisty?

The problem seems to be the patch 07_azureus-themed.dpatch. If you pull this patch out of the ubuntu source, azureus will not crash at startup (unless you are experiencing some issue that is different from mine)

New diff:

* I should not remove the patch, but only disable it in 00list.

* Also, I have of course verified that azureus also works correctly and that you can upload/download stuff etc with this modification.

New diff:

* The version number should be ubuntu1.1
* The distribution should be feisty-proposed

New diff:

* This is still proposed1

(sorry for spamming)

I have attached the log file from azureus that identifies the particular startup crash that my debdiff (removal of the azureus-themed patch) solves since the bug reports above suggest that there may be several distinct causes.

An effort should be made to distinguish them instead of just marking all the bugs with similar effect as duplicates of this bug.

For instance my debdiff above solves the startup issue on both my desktop and my laptop that are running feisty, but a friend of mine who also has startup problems reported that his azureus still crashed after installing my patched azureus package.

Further to my comments in bug 74799, these bugs are not dups of each other. All the other dups appear correct except for bug 83780, which, like bug 74799, is a dup of bug 68020. Can you confirm that your friend in fact has this second problem?

crimsun@FUN:/tmp$ debdiff azureus_2.5.0.0repack1-0ubuntu1.dsc azureus_2.5.0.0repack1-0ubuntu1.1~proposed1.dsc
diff -u azureus-2.5.0.0repack1/debian/changelog azureus-2.5.0.0repack1/debian/changelog
--- azureus-2.5.0.0repack1/debian/changelog
+++ azureus-2.5.0.0repack1/debian/changelog
@@ -1,3 +1,14 @@
+azureus (2.5.0.0repack1-0ubuntu1.1~proposed1) feisty-proposed; urgency=low
+
+ [ Ulf Karlsson ]
+ * Don't apply debian/patches/07_azureus-themed.dpatch since it
+ causes a segmentation fault at startup. LP: #57875.
+
+ [ Daniel T. Chen ]
+ * debian/control: Adhere to DebianMaintainerField.
+
+ -- Daniel T Chen <email address hidden> Sat, 28 Apr 2007 21:34:27 -0400
+
 azureus (2.5.0.0repack1-0ubuntu1) feisty; urgency=low

   [ John Dong ]
diff -u azureus-2.5.0.0repack1/debian/control azureus-2.5.0.0repack1/debian/control
--- azureus-2.5.0.0repack1/debian/control
+++ azureus-2.5.0.0repack1/debian/control
@@ -1,7 +1,8 @@
 Source: azureus
 Section: net
 Priority: optional
-Maintainer: Shaun Jackman <email address hidden>
+Maintainer: Ubuntu MOTU Developers <email address hidden>
+XSBC-Original-Maintainer: Shaun Jackman <email address hidden>
 Build-Depends: cdbs, libbcprov-java (>= 1.33), libgnucrypto-java, libcommons-cli-java, liblog4j1.2-java, libseda-java, libswt3.2-gtk-java, libgtk-java (>= 2.9.3), junit, java-gcj-compat-dev (>= 1.0.65-6), ant, debhelper (>> 4.0.0), sharutils, dpatch
 Standards-Version: 3.7.2

diff -u azureus-2.5.0.0repack1/debian/patches/00list azureus-2.5.0.0repack1/debian/patches/00list
--- azureus-2.5.0.0repack1/debian/patches/00list
+++ azureus-2.5.0.0repack1/debian/patches/00list
@@ -2,7 +2,6 @@
 01_azureus-remove-win32-PlatformManagerUpdateChecker
 02_azureus-cache-size
 03_azureus-remove-manifest-classpath
-07_azureus-themed
 08_azureus-rh-bugzilla-180418
 09_azureus-no-shared-plugins
 12_azureus-no-updates-PluginInitializer

The above comments are getting quite complex.

It seems that there is an issue with /usr/share/java/Azureus2.jar, which is simply fixed by overwriting it with Azureus2.5.0.0.jar from http://prdownloads.sourceforge.net/azureus/Azureus2.5.0.0.jar?download (a big thank-you to Juho Lehto, who posted above on 2006-10-16)

I've had no problems with this approach, under Kubuntu Fiesty.

Indeed, getting Azureus from SourceForge often seems to resolve problems.

However, there is also confusion amongst different bugs, which I have being trying to clean up. "It doesn't start" is not particularly helpful and confuses separate causes. It is unclear whether the SourceForge version works because Ubuntu patches are causing problems or whether it is choice of compiler or both. It sounds like Gary Coady (on 2006-10-09) is closest to locating the cause here with his fourth point: lack of GTK initialisation. I'm not sure why this problem doesn't affect everyone - maybe also choice of JRE is an issue.

Accepted into feisty-proposed, please go ahead with QA testing.

Works well for me so far, nice work.

2.5.0.0repack1-0ubuntu1.1~proposed1

ons 2007-05-02 klockan 14:24 +0000 skrev Martin Pitt:
> Accepted into feisty-proposed, please go ahead with QA testing.
>
> ** Tags added: verification-motu-needed
>
> ** Changed in: azureus (Ubuntu)
> Status: In Progress => Fix Committed
>
--
<email address hidden>

Plus ça change, plus c'est la même chose

On 5/2/07, Martin Pitt <email address hidden> wrote:
>
> Accepted into feisty-proposed, please go ahead with QA testing.
>

I've installed azureus_2.5.0.0repack1-0ubuntu1.1~proposed1_all.deb (from
feisty-proposed) and it's running well, using sun-java5 (1.5.0-08-0ubuntu1).

-james.

Working for me with Sun Java 6.

Im havign the same problem on ubuntu gutsy i386 with sun java 6.
Commandline output:
alec@alecomp:~/Videos/Originals$ azureus

***MEMORY-WARNING***: [1800]: GSlice: g_thread_init() must be called before all other GLib functions; memory corruption due to late invocation of g_thread_init() has been detected; this program is likely to crash, leak or unexpectedly abort soon...
/home/alec/.themes/Clearlooks_blackblue/gtk-2.0/gtkrc:53: Clearlooks configuration option "sunkenmenu" is not supported and will be ignored.
DEBUG::Thu May 31 19:09:14 BST 2007::org.gudy.azureus2.ui.swt.mainwindow.Initializer::run::314:
  java.lang.UnsupportedClassVersionError: org/gnu/glib/GObject (Unsupported major.minor version 49.0)
        at java.lang.ClassLoader.defineClass0(Native Method)
        at java.lang.ClassLoader.defineClass(ClassLoader.java:539)
        at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:123)
        at java.net.URLClassLoader.defineClass(URLClassLoader.java:251)
        at java.net.URLClassLoader.access$100(URLClassLoader.java:55)
        at java.net.URLClassLoader$1.run(URLClassLoader.java:194)
        at java.security.AccessController.doPrivileged(Native Method)
        at java.net.URLClassLoader.findClass(URLClassLoader.java:187)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:289)
        at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:274)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:235)
        at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:302)
        at java.lang.ClassLoader.defineClass0(Native Method)
        at java.lang.ClassLoader.defineClass(ClassLoader.java:539)
        at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:123)
        at java.net.URLClassLoader.defineClass(URLClassLoader.java:251)
        at java.net.URLClassLoader.access$100(URLClassLoader.java:55)
        at java.net.URLClassLoader$1.run(URLClassLoader.java:194)
        at java.security.AccessController.doPrivileged(Native Method)
        at java.net.URLClassLoader.findClass(URLClassLoader.java:187)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:289)
        at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:274)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:235)
        at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:302)
        at org.gudy.azureus2.ui.swt.ImageRepository.getThemedIcon(ImageRepository.java:73)
        at org.gudy.azureus2.ui.swt.ImageRepository.loadImages(ImageRepository.java:130)
        at org.gudy.azureus2.ui.swt.mainwindow.Initializer.run(Initializer.java:214)
        at org.gudy.azureus2.ui.swt.mainwindow.SWTThread$1.runSupport(SWTThread.java:115)
        at org.gudy.azureus2.core3.util.AERunnable.run(AERunnable.java:38)
        at java.lang.Thread.run(Thread.java:534)

I've attached a screenshot of the crashed splash screen

My friend is having the same problem on gutsy too.

I tried it 3 times. On the 1st and 3rd, it popped up the language select screen for about a second. On the 2nd, it crashed on the splash screen, and I ctrl+c'ed it. I've attached a screenshot.
Terminal output:
robin@robin-desktop:~$ azureus
DEBUG::Sat Jun 02 14:40:36 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
DEBUG::Sat Jun 02 14:40:37 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,CryptoManagerImpl::<init>::-1,CryptoManagerImpl::getSingleton::-1,CryptoManagerFactory::getSingleton::-1,AzureusCoreImpl::<init>::-1,AzureusCoreImpl::create::-1,AzureusCoreFactory::create::-1,Main::<init>::-1,Main::main::-1

***MEMORY-WARNING***: [9078]: GSlice: g_thread_init() must be called before all other GLib functions; memory corruption due to late invocation of g_thread_init() has been detected; this program is likely to crash, leak or unexpectedly abort soon...
Aborted (core dumped)

robin@robin-desktop:~$ azureus
DEBUG::Sat Jun 02 14:41:35 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
DEBUG::Sat Jun 02 14:41:35 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,CryptoManagerImpl::<init>::-1,CryptoManagerImpl::getSingleton::-1,CryptoManagerFactory::getSingleton::-1,AzureusCoreImpl::<init>::-1,AzureusCoreImpl::create::-1,AzureusCoreFactory::create::-1,Main::<init>::-1,Main::main::-1

***MEMORY-WARNING***: [9180]: GSlice: g_thread_init() must be called before all other GLib functions; memory corruption due to late invocation of g_thread_init() has been detected; this program is likely to crash, leak or unexpectedly abort soon...
robin@robin-desktop:~$ azureus
DEBUG::Sat Jun 02 14:42:35 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider available
    SESecurityManager::initialise::-1,ConfigurationChecker::setSystemProperties::-1,ConfigurationManager::initialise::-1,ConfigurationManager::getInstance::-1,LoggerImpl::init::-1,Logger::<clinit>::-1,Class::initializeClass::-1,Logger::isEnabled::-1,StartServer::<init>::-1,Main::<init>::-1,Main::main::-1
DEBUG::Sat Jun 02 14:42:35 GMT+01:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::-1:
  No SSL provider av...

I've unsubscribed ubuntu-universe-sponsors, as it appears tha the necessary upload is complete. Please resubscribe if I am in error, and additional sponsorship is required. Thank you.

   azureus | 2.5.0.0repack1-0ubuntu1.1~proposed1 | feisty-proposed/universe | source, all

Daniel, you need a manual upload to -updates and drop the ~proposed1 suffix. I do not want to copy-package this, since the suffix is confusing.

You also need to fix this in gutsy before it goes into a stable update.

Are we sure that those gutsy crashes are the same bug? They don't look like the same crash report and neither screenshot shows the splash stuch at "loading images" as described here or in the duplicates. Perhaps a separate bug should be filed for this gutsy problem.

Also, is the patch simply disabling themes for Azureus? Isn't that more of a workaround than a solution?

DEBUG::Wed Jul 25 13:32:39 GMT+01:00 2007::com.aelitis.azureus.core.networkmanager.impl.tcp.VirtualChannelSelectorImpl::select::528:
  VirtualChannelSelector: No progress for op 1: listener = class com.aelitis.azureus.core.clientmessageservice.impl.NonBlockingReadWriteService$2, count = 10, socket: open = true, connected = true
    VirtualChannelSelector::select::272,NonBlockingReadWriteService$1::runSupport::79,AEThread::run::69
Aborted (core dumped)

Mine loaded fine the first time, now it crashes on startup with no promp for bug report, so here is the dump.

I'm not sure if this is the same issue, but Azureus crashes on boot for me as well. I have attached a log and the core dump.

Matt, I haven't noticed any other class cast exceptions like your log
showed, but I might have just missed them.

I noticed from the log that you're running Java 1.6.0_02-b05 - does that
mean you're on gutsy (ubuntu 7.10)? In general the more information you can
give about your software versions and how they were installed the more
helpful your report will be.

Regards,

-james.

I am running Gutsy and the latest (though not necessarily stable) versions of everything. I uninstalled and reinstalled Azureus from the repos, but it still won't work. Should I set up a new error report for this?

~Matt

Why don't update azureus to 2.5.0.4 in Gutsy? azureus 2.5.0.4 is the latest version which is very stable and also works great with Ubuntu Gutsy.

From now on, replacing the /usr/share/java/Azureus2.jar

with the jar file that can be downloaded here :

http://easynews.dl.sourceforge.net/sourceforge/azureus/Azureus2.5.0.4.jar

fix the problem.

Just wanted to confirm Mertiki's work-around... runs fine here on Gutsy with latest updates.

Can someone please give an update about this bug? It is verification-done, but comment 82 leaves some doubts. In addition, somone must actually do a no-change upload without ~proposed to -updates.

Also, I will not accept anything into -updates which is not fixed in gutsy yet.

I think it would help if people posted the hotspot crash reports (and descriptions, Shane) not just the output. Perhaps this should be made clear in all Ubuntu bug reporting guides? It would be great if this could be an automatic warning for people trying to report against Java-based packages.

(additionally, posts about upstream working are not very useful as it seems to be Ubuntu/packaging issue at the moment)

Also, to clarify: Martin, what makes you think this isn't fixed in Gutsy? Has the change been incorporated into any repo? Comment #79 says that downgrading to feisty-proposed fixed it, in case you hadn't seen the clipped portion. I'm Internet-less on my Ubuntu machine and have been for weeks, so I can't really play with this.

I'm running Gutsy with Java6 on 32 bit Intel and it crashes at startup for me. What specifics would you like?

I'm having the same problem in edgy. I've found that azureus starts and runs correctly whenever i DELETE the ~/.azureus/logs folder

Running Gutsy tribe 5 with sun-java6-bin and azureus now doesn't crash. I tested a lot of things and also downloaded a file from vuze with azureus and I don't get it to crash.

It's pretty weird because apparently, there has been no azureus update, maybe a sun-java6-bin update?

I installed azureus on several Ubuntu version and always with the same paquets, it's the first time that azureus doesn't crash on startup.

In the same way, it's the first time that I did never see a alert sidebar in azureus, which seems to be the cause of the crash. Can someone look if azureus works great in Ubuntu Gutsy now?

Azureus finally crashed after a reboot, and it's now impossible to make it starts without crashing.

Did you tried deleting the "logs" directory?

with the lastest azureus (3.0.2.0) installed manually over the sypnatic version doesnt crashes with JRE

Sebastian : I didn't remove the logs directory, but removing the warnings will not correct the bug, it will hide it.

I confirm, the 3.0.2.2 azureus release works great with Ubuntu too. Why azureus isn't just upgraded to the latest version before it's too late for gutsy? I think it would close a lot of problems. The source package wouldn't be hard to upgrade, no modification has been made in the source files from the orig.tar.gz package.

Sebastian : Does azureus stop permanently crashing if you delete the logs file?

2007/9/11, Saïvann Carignan <email address hidden>:
> Sebastian : Does azureus stop permanently crashing if you delete the
> logs file?
>

it stops until there's another entry in the log... I agree with you,
removing the directory will not correct the bug, but it may help to
find where the bug is..

Having same issue with Gutsy Tribe one. Attached is the terminal output just before crashing. I then delete the logs directory and all works. I will attach log files if needed.

Correction on previous post: I am using Gutsy Tribe 5

I tried replacing Azureus2.jar with Azureus2.5.0.4, as suggested above, still get error message:

changeLocale: *Default Language* != English (United States). Searching without country..
changeLocale: Searching for language English in *any* country..
changeLocale: no message properties for Locale 'English (United States)' (en_US), using 'English (default)'
Exception in thread "main" java.lang.UnsupportedClassVersionError: org/eclipse/swt/widgets/Display (Unsupported major.minor version 49.0)
        at java.lang.ClassLoader.defineClass0(Native Method)
        at java.lang.ClassLoader.defineClass(ClassLoader.java:539)
        at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:123)
        at java.net.URLClassLoader.defineClass(URLClassLoader.java:251)
        at java.net.URLClassLoader.access$100(URLClassLoader.java:55)
        at java.net.URLClassLoader$1.run(URLClassLoader.java:194)
        at java.security.AccessController.doPrivileged(Native Method)
        at java.net.URLClassLoader.findClass(URLClassLoader.java:187)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:289)
        at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:274)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:235)
        at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:302)
        at org.gudy.azureus2.ui.swt.Main.<init>(Main.java:80)
        at org.gudy.azureus2.ui.swt.Main.main(Main.java:180)

I guess I have a similar problem ...

vonkad@eg0022:~$ azureus
DEBUG::Tue Sep 25 13:40:29 GMT+02:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::140:
  No SSL provider available
    SESecurityManager::initialise::52,ConfigurationChecker::setSystemProperties::145,ConfigurationManager::initialise::138,ConfigurationManager::getInstance::71,LoggerImpl::init::90,Logger::<clinit>::48,Class::initializeClass::-1,StartServer::<init>::70,Main::<init>::56,Main::main::162
changeLocale: *Default Language* != en (US). Searching without country..
changeLocale: Searching for language en in *any* country..
changeLocale: no message properties for Locale 'en (US)' (en_US), using 'English (default)'
DEBUG::Tue Sep 25 13:40:31 GMT+02:00 2007::org.gudy.azureus2.core3.security.impl.SESecurityManagerImpl::initialise::140:
  No SSL provider available
    SESecurityManager::initialise::52,CryptoManagerImpl::<init>::73,CryptoManagerImpl::getSingleton::60,CryptoManagerFactory::getSingleton::33,AzureusCoreImpl::<init>::155,AzureusCoreImpl::create::92,AzureusCoreFactory::create::46,Main::<init>::143,Main::main::162
DEBUG::Tue Sep 25 13:40:34 GMT+02:00 2007::com.aelitis.azureus.core.networkmanager.impl.tcp.VirtualChannelSelectorImpl::select::528:
  VirtualChannelSelector: No progress for op 1: listener = class com.aelitis.azureus.core.clientmessageservice.impl.NonBlockingReadWriteService$2, count = 10, socket: open = true, connected = true
    VirtualChannelSelector::select::272,NonBlockingReadWriteService$1::runSupport::79,AEThread::run::69
Aborted (core dumped)
vonkad@eg0022:~$

I have this bug also, doesn't seem to respond to deleting the .azureus/log file

Fresh updated Gutsy

Exception in thread "main" java.lang.UnsupportedClassVersionError: org/eclipse/swt/widgets/Display (Unsupported major.minor version 49.0)
        at java.lang.ClassLoader.defineClass0(Native Method)
        at java.lang.ClassLoader.defineClass(ClassLoader.java:539)
        at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:123)
        at java.net.URLClassLoader.defineClass(URLClassLoader.java:251)
        at java.net.URLClassLoader.access$100(URLClassLoader.java:55)
        at java.net.URLClassLoader$1.run(URLClassLoader.java:194)
        at java.security.AccessController.doPrivileged(Native Method)
        at java.net.URLClassLoader.findClass(URLClassLoader.java:187)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:289)
        at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:274)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:235)
        at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:302)
        at org.gudy.azureus2.ui.swt.Main.<init>(Main.java:147)
        at org.gudy.azureus2.ui.swt.Main.main(Main.java:162)

Child

This is a "ME TOO" post. Though someone asked for HotSpot error reports, so I'm attaching mine here. I hope this helps.

I'm having similar issues, they are temporarily resolved by deleting "~/.azureus/logs/". I'm running Feisty. I'm running IA32, java6.

From running azureus:
-----
DEBUG::Sat Sep 29 22:15:36 PDT 2007::com.aelitis.azureus.core.networkmanager.VirtualChannelSelector::select::272:
  Caught exception on selector.select() op: Operation not permitted
    NonBlockingReadWriteService$1::runSupport::80,AEThread::run::69
java.io.IOException: Operation not permitted
        at sun.nio.ch.EPollArrayWrapper.epollCtl(Native Method)
        at sun.nio.ch.EPollArrayWrapper.updateRegistrations(EPollArrayWrapper.java:202)
        at sun.nio.ch.EPollArrayWrapper.poll(EPollArrayWrapper.java:183)
        at sun.nio.ch.EPollSelectorImpl.doSelect(EPollSelectorImpl.java:65)
        at sun.nio.ch.SelectorImpl.lockAndDoSelect(SelectorImpl.java:69)
        at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:80)
        at com.aelitis.azureus.core.networkmanager.impl.tcp.VirtualChannelSelectorImpl.select(VirtualChannelSelectorImpl.java:446)
        at com.aelitis.azureus.core.networkmanager.VirtualChannelSelector.select(VirtualChannelSelector.java:272)
        at com.aelitis.azureus.core.clientmessageservice.impl.NonBlockingReadWriteService$1.runSupport(NonBlockingReadWriteService.java:80)
        at org.gudy.azureus2.core3.util.AEThread.run(AEThread.java:69)
#
# An unexpected error has been detected by Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00000000, pid=31522, tid=3084737424
#
# Java VM: Java HotSpot(TM) Client VM (1.6.0-b105 mixed mode, sharing)
# Problematic frame:
# C 0x00000000
#
# An error report file with more information is saved as hs_err_pid31522.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)
-----
Attached is the HotSpot error report.

I had the same problem but downloaded the latest version of Azureus and replaced the Azureus2.jar file (attached) in /usr/share/java and it seems to be working permanently.

This is a me too on the bug (same symptoms) though the root cause seems different. This is what I get when running azureus on a default gutsy system:

> azureus
changeLocale: *Default Language* != English (United States). Searching without country..
changeLocale: Searching for language English in *any* country..
changeLocale: no message properties for Locale 'English (United States)' (en_US), using 'English (default)'
should load blocklist from: http://www.bluetack.co.uk/config/spconfig.txt
Retrieving data from: http://www.bluetack.co.uk/config/splist.zip
#
# An unexpected error has been detected by Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x8f79474b, pid=10382, tid=3085114256
#
# Java VM: Java HotSpot(TM) Server VM (1.6.0_03-b05 mixed mode)
# Problematic frame:
# C [libglibjni-0.4.so+0xb74b]
#
# An error report file with more information is saved as hs_err_pid10382.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

Upgrading the Azureus2.jar as listed in the last post does get me around the issue.

Confirmed on latest gutsy amd64

Also confirmed a java virtual machine crash with latest gutsy amd64 and Java 6. Installing Azureus from the sourceforge fixed the issue, judging from the error messages the problem might be with the jni libraries or interfacing with them.

Azureus should work fine but if you want do display more informations (right click more details) azureus crash :

$ azureus
#
# An unexpected error has been detected by Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x8f7b174b, pid=6636, tid=3084192656
#
# Java VM: Java HotSpot(TM) Server VM (1.6.0_03-b05 mixed mode)
# Problematic frame:
# C [libglibjni-0.4.so+0xb74b]
#
# An error report file with more information is saved as hs_err_pid6636.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

I run the gutsy version

I can also confirm this bug on Gutsy. Right after startup, the main window shows for a split second, then

#
# An unexpected error has been detected by Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x062145b9, pid=8059, tid=3084987280
#
# Java VM: Java HotSpot(TM) Client VM (1.6.0_03-b05 mixed mode, sharing)
# Problematic frame:
# V [libjvm.so+0x2145b9]
#
# An error report file with more information is saved as hs_err_pid8059.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#
Aborted (core dumped)

I can confirm this bug, but when I ran Azureus and it went through the first-time setup wizard, it worked fine (I ran it under sudo su).

I reinstalled both azureus packages in synaptic. After getting through the configuration wizard--it works just fine again.

In my case I got my Azureus working again with running the first-time setup wizard once or twice. But after restart it started crashing again later on. So please monitor the situation for a while and let us know if you got rid of it permanently or just temporarily?

I got the same problem on official 7.10 release on AMD64.

#
# An unexpected error has been detected by Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00002aaac39bab4a, pid=5426, tid=1076017488
#
# Java VM: Java HotSpot(TM) 64-Bit Server VM (1.6.0_03-b05 mixed mode)
# Problematic frame:
# C [libglibjni-0.4.so+0xcb4a]
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
#

GREAT SUCCESS! Long story short: I've got a working solution for this bug that can be uploaded to Hardy for certain, backported to Gutsy for certain, and may be candidate for SRU to Gutsy too.

Long story:
Our current Azureus 2.5.0.0 package uses a heavily customized build environment deriving from Fedora 6 packaging, and contains several patches specifically to make Azureus compile with GCJ. It would seem like a combination of those patches, and the act of compiling with GCJ, produces broken class files that not even Sun Java can interpret.

Meanwhile, an open source Icedtea (Java7) JVM is now available which works fine with Azureus. Debian also has very clean packaging for Azureus which works beautifully on Ubuntu with icedtea with a few minor changes.

What I have done to get working Azureus:

(1) Base off Debian azureus 2.5.0.4-1, discarding all Ubuntu changes
(2) Patch the makefile, azureus launcher, and build dependencies with Ubuntu SWT locations
(3) Patch the build and runtime dependencies to suggest building/running off icedtea
(4) Patch the azureus launcher to explicitly call the SWT interface; the decider logic crashes.

END RESULT, on Gutsy:

Building with GCJ:
   ALL RUNTIMES FAIL: Infinite deadlock with GCJ, crashes with Sun/Icedtea
Building with Icedtea:
   GCJ: SUCCESS: Runs, a bit slowly but not bad at all
   Icedtea, sun-java6, sun-java5: SUCESS: Runs smoothly and very fast

I am preparing a few packages for upload to hardy, and for testing.

Hello,

This is what worked for me. ( Env: Ubuntu 7.10)

1.- Select the proper Java run time by running
      sudo update-alternatives --config java

output:
Selection Alternative
-----------------------------------------------
          1 /usr/bin/gij-4.2
+ 2 /usr/lib/j2se/1.4/bin/java
* 3 /usr/lib/jvm/java-1.5.0-sun/jre/bin/java
          4 /usr/lib/jvm/java-6-sun/jre/bin/java

( I entered 2 at the prompt, previous option was 3)

Now my java version reads "1.5.0_13"

2.- I downloaded Azureus_3_0_linux.tar and extracted Azureus2.jar to /usr/share/java/
     (make a backup of Azureus2.jar just in case)

3.- started Azureus

This works for me, but it may not work for others...

Cheers

This is a testing .deb for Gutsy built using a pbuilder. Please confirm on your JVM of choice. Should work with gcj, java5, java6, and icedtea-java7-jre

Feedback is appreciated and essential to timely fix :)

DISCLAIMER:
This is a testing package, do not use on production systems. I have made my best effort of packaging this safely to all Ubuntu guidelines and standards, and made very minimal changes to Debian packaging. Provided that this upload goes into Hardy, it should not conflict with any future upgrades.

Marking Feisty as Incomplete -- have no idea how to build with a icedtea JVM in anything but Gutsy and higher, excluding Backports.

This can be SRU'ed to Gutsy, by uploading the 2.5.0.4 debian version of the package + ubuntu diff.gz, with a version number tweak to be less than Hardy. I know it's extremely uncustomary to update to a new upstream version but I assure the motu-sru folks that this is better than what we currently have, and any other solutions for the following reasons:

(1) Current packaging is a mess: It's a repacked tarball with a COMPLETELY different directory structure from upstream. Debian package has identical to upstream directory structure and will thus be easier to maintain in the future.
(2) Current packaging uses a GCJ-specific buildsystem. It is not obvious to me how to replace the compiler with icedtea without resorting to ripping out debian/rules completely, or doing some sort of awful hackjob. Remember that building Azureus with GCJ = crash across the board.
(3) Current packaging applies a few patches that are specific to making GCJ work. I have no idea what kind of negative consequences that may have on compiling with icedtea.
(4) Current packaging does not work for anyone, under any JVM, because GCJ compiling it breaks the resulting bytecode. This new package can only be better, not worse. Azureus has been neglected the whole Gutsy release cycle and the package has not changed since Feisty, when I last touched it.
(5) The 2.5.0.4-1 Debian package has been promoted to Debian Testing for over 3 months without any bug reports against it (other than people wanting 3.0.0.0, unrelated to QA status of this package). I have only made trivial modifications to the Debian package so I firmly believe this package is high quality enough for a stable Ubuntu release.

I would be thrilled to be the "contact" for this bug, and Azureus in general, and promise to guide this update through, answer any questions about it, and take responsibility for any possible regressions. Please consider these arguments for a SRU under these unusual circumstances.

jdong-

I looked over this debdiff, and you will need to address the version number for the SRU properly as well as the target to be gutsy-proposed.

I haven't tested the resultant build myself yet, but as soon as a member from motu-uvf acks this on the bug, i'll be glad to help you get this into proposed.

I'm totally aware that the current version number is not compliant with SRU policies, but currently that's the last of my concerns. My #1 worry is that motu-uvf will refuse to ack a new upstream version into a stable Ubuntu release, and I'd like to get a green-light on that before preparing a SRU. The current package is in anticipation of Hardy opening for uploads.

I'm using kubuntu desktop 7.10 and now want to download ubuntu server 7.10 with azureus and I've got the same problem, azureus is crashing. If I delete logs-dir, it starts again, but If I double click on the ubuntu torrent to view details, it crashes again and I have to delete logs/ to get starting it again.

Can you please try the .deb I attached to this bug report? (Set your
default Java to a Sun or icedtea Java implementation for optimal speed,
but it should work with any Java)

I need success reports before I can continue pursuing this update.

On Mon, Oct 22, 2007 at 03:11:55PM -0000, Kev wrote:
> I'm using kubuntu desktop 7.10 and now want to download ubuntu server
> 7.10 with azureus and I've got the same problem, azureus is crashing. If
> I delete logs-dir, it starts again, but If I double click on the ubuntu
> torrent to view details, it crashes again and I have to delete logs/ to
> get starting it again.

Hi John,

Thank you very much for sorting this out!

Regarding feisty, to the best of my knowledge, everyone who has tried the 2.5.0.0repack1-0ubuntu1.1~proposed1 deb has reported success, both with feisty and edgy. But the package needs to be renamed in order to make it into the archives. (I'm not sure at all what build magic made this deb work.)

Another idea is that feisty included the eclipse compiler (ecj) package, which is well-regarded for its capabilities. Can it be used successfully (in place of the nonexistent icedtea javac) to build new packages in feisty?

-james.

i've been having this same problem. the azureus_2.5.0.4-1ubuntu1~0jdong1~7.10.gutsy1_all.deb seems to start (using GCJ) but it threw up error messages:
SWT library loaded from "/usr/lib/java" can't be automatically updated from version 3236 to 3318 (must be loaded from "/home/shawn/.azureus"). please see the wiki for details

when i try to download the gutsy dvd torrent i click it and open with azureus it pops up this erreo message:
'/tmp/gutsy-dvd-i386.iso.torrent' could not be opened:

Error reading torrent file '/home/shawn/Torrents/gutsy-dvd-i386.iso.torrent - Failed to calculate hash: NullPointerException
and it wont download it

if i click and drag the torrent file from desktop to azureus window it says:
Piece count: 3
Hashing files
TOTorrentFileHasher: file read fails 'java.lang.NullPointerException', ShareResource: Torrent create failed

so it starts but doesn't download

p.s. i'm using gutsy i386

Well, here's the deal:

(1) IF: I get word from an archive admin and motu-uvf that a SRU consisting of
entirely new packages is acceptable, then fine, I will look into finding all
working Java stacks for Feisty.

(2) IF I do not (which I think is more likely): I will handle it as a backport to gutsy (trivial), and fiesty and perhaps edgy ( which would likely just require an icedtea-java7-* backport, which should also not take too much effort )

If anyone out there would like to test the 2.5.0.4-1ubuntu1 package sources
compiling against a different Java stack in the repositories, they are welcome
to do so.

(1) The updater should be disabled with a clean ~/.azureus directory, I cannot
stop the updater jars in your profile from trying to update Azureus's SWT
libraries.

(2) The torrent check thing... can you try again with either sun Java or
icedtea-java7-jre?

i deleted the ~/.azureus directory and it still pops up the same error message:
SWT library loaded from "/usr/lib/java" can't be automatically updated from version 3236 to 3318 (must be loaded from "/home/shawn/.azureus"). please see the wiki for details

but now i can download torrents
still using the same java

Thank you for the update,

I have been able to confirm this error when I hit Check For Updates on
the menu. Looks like the Debian patches for removing the update checker
are incomplete. I will investigate how to disable this error message.

Note that this does not affect functionality, and should only pop up
on during the periodic startup update check, and the error disappears
after a few seconds on its own, a la libnotify-bubble. At any rate, it
is an annoyance, but not a show-stopper to this fix.

On Tue, Oct 23, 2007 at 12:01:29AM -0000, Shawn vega wrote:
> i deleted the ~/.azureus directory and it still pops up the same error message:
> SWT library loaded from "/usr/lib/java" can't be automatically updated from version 3236 to 3318 (must be loaded from "/home/shawn/.azureus"). please see the wiki for details
>
>

Hi Shawn,

Just an update, I have fixed the said annoyance by disabling the SWT check in the source. I cleaned out my Azureus tree and only got an update for azupdater, which applied successfully. Subsequent startups did not have nagging updates or errors.

The change is currently in bzr on Launchpad; along with another one that enforces the usage of a Icedtea or Sun Java 5/6 stack. (Turns out GCJ is EXTREMELY slow hash-checking to the point that it's unacceptable). I will not be rolling out a testing deb until I get some more feedback on the changes, to avoid spamming with a huge storm of updated debs and bloated version numbers.

I'm a little more confused now. I am having the same problem as Kev. I don't see where to download the fixed .deb on launchpad. Is this going to eventually make it to the repositories?

I'm a novice so layman's terms would be appreciated :)

> Can you please try the .deb I attached to this bug report? (Set your
> default Java to a Sun or icedtea Java implementation for optimal speed,
> but it should work with any Java)
>
> I need success reports before I can continue pursuing this update.
>
> On Mon, Oct 22, 2007 at 03:11:55PM -0000, Kev wrote:
>> I'm using kubuntu desktop 7.10 and now want to download ubuntu server
>> 7.10 with azureus and I've got the same problem, azureus is crashing. If
>> I delete logs-dir, it starts again, but If I double click on the ubuntu
>> torrent to view details, it crashes again and I have to delete logs/ to
>> get starting it again.

http://launchpadlibrarian.net/10110296/azureus_2.5.0.4-1ubuntu1%7E0jdong1%7E7.10.gutsy1_all.deb

Here you go :) Linked on the left-side bar but sometimes hard to find.

Yay, it works! Thanks a ton!

I am looking forward to the "SWT library loaded..." bug-fixed version whenever it's convenient.

ok it starts and everything, adds torrents, adds peers, but it doesn't actually download any files. i'm going to install a different java and see if that changes anything.

Shawn,

Please use icedtea-java7-jre from Universe, (or you can use the
sun-java*-jre flavors from multiverse if you don't particularly
subscribe to a FOSS-vegan diet)

It is the only JRE that I am supporting for the new packaging to
azureus, and the next revision to the package will enforce that as a
runtime dependency.

On Tue, Oct 23, 2007 at 03:14:21AM -0000, Shawn vega wrote:
> ok it starts and everything, adds torrents, adds peers, but it doesn't
> actually download any files. i'm going to install a different java and
> see if that changes anything.
>

ok everything works great with icetea-java7-jre

Installed Gutsy, Installed Java 6, Installed Azureus, Ran Azureus -- CRASH!

Installed test .deb above, it works again. So looks like a winner to me.

By the way, anyone else getting weird 'NotaFile' type errors when trying to open torrents with Right-Click Azureus? For some reason the only way I can open torrents right now is via the open torrent dialog inside Azureus, which is -very- annoying.

The attached debdiff (http://launchpadlibrarian.net/10110278/debdiff) looks reasonable to me. So I propose the following:

 * Get this uploaded into Hardy, built, and test that it now works OOTB.
 * Create an automated backport to gutsy-backports and test it.
 * Once it is confirmed to work, copy the gusy-backports source and binary into gutsy-updates.

This approach should provide step-by-step verification without causing any unnecessary manual work. What do you think?

Wrt. SRU policy, since the gutsy package is not working at all, there's little that can be broken any further, so I think it is resonable to do more invasive changes in this case.

Martin, I fully agree your points here. Gonna coordinate with John about it and get it done as soon as possible. Thanks.

I also fully agree with you, Martin. Thank you for your time, and thanks everyone subscribed to this bug for your continued patience as we sort through this and put this bug to rest!

GUTSY TESTERS:

Here is a new testing deb that should resolve all the issues with the previous one:
http://jdong.mit.edu/~jdong/motu/azureus_2.5.0.4-1ubuntu2~gutsy0.1_all.deb

---- changelog ----
  * Makes previous upload functional without changing default
    /usr/bin/java alternative (LP: #57875)
  * Tighten build and runtime dependencies -- GCJ doesn't work
    at all (not even as runtime; slow hash-checking, no transfers)
  * Correct Original-Maintainer field name.
  * Modify azureus script to detect and choose a optimal JRE.
    First try icedtea, then look for java6 or java5 from Sun, then
    fall back to /usr/bin/java symlink after complaining
  * Strip out the SWT update on start (caused annoying error bubble
     about not being able to write to /usr/lib/java on first startup)

 -- John Dong <email address hidden> Mon, 22 Oct 2007 21:23:24 -0400
----------------------

Testing would be appreciated. If people can also test for me with an *emptied* ~/.azureus, that'd be great too. I've received a report that "azupdater" wants to update twice (update once, restart, it wants to update AGAIN, then update again, and it works).

I'd also like opinions from everyone if I should disable all update checking at startup, not just SWT.

So far so good on the last .deb. If problems occur later I will update.

PHASE 1:

ubuntu-universe-sponsors, please upload http://jdong.mit.edu/~jdong/motu/azureus_2.5.0.4-1ubuntu2.dsc to Hardy.

Hi,

On my 32bit Gutsy everything seems to work all right. I removed
.azureus/ before installing the package. azupdater is indeed installed
twice.

Thanks,
David Vonka

John Dong wrote:
> GUTSY TESTERS:
>
> Here is a new testing deb that should resolve all the issues with the previous one:
> http://jdong.mit.edu/~jdong/motu/azureus_2.5.0.4-1ubuntu2~gutsy0.1_all.deb
>
> ---- changelog ----
> * Makes previous upload functional without changing default
> /usr/bin/java alternative (LP: #57875)
> * Tighten build and runtime dependencies -- GCJ doesn't work
> at all (not even as runtime; slow hash-checking, no transfers)
> * Correct Original-Maintainer field name.
> * Modify azureus script to detect and choose a optimal JRE.
> First try icedtea, then look for java6 or java5 from Sun, then
> fall back to /usr/bin/java symlink after complaining
> * Strip out the SWT update on start (caused annoying error bubble
> about not being able to write to /usr/lib/java on first startup)
>
> -- John Dong <email address hidden> Mon, 22 Oct 2007 21:23:24 -0400
> ----------------------
>
> Testing would be appreciated. If people can also test for me with an
> *emptied* ~/.azureus, that'd be great too. I've received a report that
> "azupdater" wants to update twice (update once, restart, it wants to
> update AGAIN, then update again, and it works).
>
> I'd also like opinions from everyone if I should disable all update
> checking at startup, not just SWT.
>

azureus (2.5.0.4-1ubuntu1) hardy; urgency=low

  * The "Ubuntu has working Azureus after 3 years" release (LP: #57875)
  * Rebase Ubuntu packaging off Debian
  * Fix launcher script and Makefile to use Ubuntu swt.jar path
  * Fix launcher script to call the correct entry point
  * Change build-deps on swt to match Ubuntu package names
  * Build with icedtea-java7 stack
  * Tighten runtime dependencies -- GCJ doesn't work
    at all (not even as runtime; slow hash-checking, no transfers)
  * Correct Original-Maintainer field name.
  * Modify azureus script to detect and choose a optimal JRE.
    First try icedtea, then look for java6 or java5 from Sun, then
    fall back to /usr/bin/java symlink after complaining
  * Strip out the SWT update on start (caused annoying error bubble
     about not being able to write to /usr/lib/java on first startup)

 -- jdong <email address hidden> Wed, 21 Oct 2007 12:34:47 -0400

[Expired for azureus (Ubuntu Feisty) because there has been no activity for 60 days.]

tor 2007-10-25 klockan 11:19 +0000 skrev Launchpad Janitor:
> [Expired for azureus (Ubuntu Feisty) because there has been no activity
> for 60 days.]

Launchpad bug?

>
--
<email address hidden>

Plus ça change, plus c'est la même chose

Yeah, Launchpad bug, please ignore.

However, at this point I cannot find a solution for Feisty (no IcedTea stack). I am still discussing this with other developers, so hang on tight. In the mean time, you can actually install the hardy or gutsy fixed debs (as soon as the stupid build daemons would build it!) on Feisty and it'll pull in a Sun Java implementation to run Azureus.

Built successfully in Hardy... I will test over the weekend in Hardy i386 and amd64 virtual machines.

Progress Update:

I uploaded an updated package into Hardy. This one implements an AZUREUS_JAVA environment variable you can set before launching Azureus, which will cause Azureus to use the JVM you specified. (For list of ideas, try update-alternatives --list java)

This is for those poor souls on AMD64 suffering the icedtea-segfault bug, or otherwise have their own JVM they'd rather run Azureus with.

As soon as this builds, we will proceed to a Gutsy backport.

TESTERS:

Need your help again: Here's a candidate gutsy-backport from Hardy 1ubuntu2 revision. It should work almost identically to the 1ubuntu1 Hardy package (i.e. everything works out of the box)

AMD64 users: icedtea crashes on your architecture (icedtea bug) -- you can use AZUREUS_JAVA to override the default (Icedtea) Java selection, for example "AZUREUS_JAVA=/usr/lib/jvm/java-6-sun/jre/bin/java azureus" as a temporary workaround.

Again, timely testing feedback will allow us to progress faster.

It worked for me on Gutsy, thank you!

+1. Seems to work on Gutsy x86 for me too.
Thanks again!

I'm using it since yesterday in my Gutsy x86 and it works fine!
thanks!! :)

Works great for me on Gutsy, +1

ACK from backporters. I know the package for 1ubuntu2 hasn't built in Hardy yet, but I've built it in pbuilder and had multiple success and not a single failure report on the testing backport. 1ubuntu1 built perfectly in Hardy and 1ubuntu2 just introduces some bash scripting changes in the shellscript launcher, no Java code changes, so I have no reason to believe there would be any build-related regressions.

John, Thanks for the updated deb it works great for me on Gutsy x86.

I still see an issue with azureus_2.5.0.4-1ubuntu2~7.10prevu1_all.deb on Gutsy: double-clicking in a torrent file gives you an error from Azureus, which is being passed an URL argument when it expects a plain file argument. The following line in /usr/share/applications/azureus.desktop needs to be fixed:

--- /usr/share/applications/azureus.desktop.original-and-buggy 2007-10-27 17:49:14.000000000 -0200
+++ /usr/share/applications/azureus.desktop 2007-11-04 22:16:43.000000000 -0200
@@ -2,7 +2,7 @@
 Encoding=UTF-8
 Categories=Java;Application;Network;FileTransfer;P2P
 Comment=peer-to-peer file distribution tool
-Exec=azureus %U
+Exec=azureus %f
 GenericName=BitTorrent client
 Icon=Azureus
 MimeType=application/x-bittorrent

works great for me too on gutsy.

1ubuntu3 uploaded to Hardy with the above .desktop file change; I figure there's not much point to making a test .deb of it (I've already spammed so many onto here :D)

Fortunately the backport hasn't been processed yet so we are still on track :)

One issue with the current candidate, though probably not unique to this version, is that it leaks memory in ways I would have thought impossible in a java app. Even if I strangle it with -Xmx32m, it will instantly gobble 50+ MB on startup, and usually bloat out to over 100 within a few minutes. I don't understand how it can use more space than is allotted to the entire JVM, but it does. Since this renders it nearly unusable on my bittorrent laptop, I've been forced to ditch Azureus. I find deluge has now improved to the point where it's better than Azureus for gnome now anyway. A pity really, since Azureus would be wonderful if it worked properly.

I cannot reproduce this happening with the standard settings on Azureus with a torrent with 100 connections. It uses about 35MB resident - shared memory on startup, and uses about 50MB with a torrent running.

Comparing with Deluge, it's roughly the same overall memory usage with active torrents, Deluge is slightly more friendly at RAM but uses more CPU power for some reason.

With that said, Azureus was not designed with extreme lightness in mind, it does use RAM to optimize performance of torrenting, and modern PC's should have more than ample resources to handle it.

I do have the SafePeer plugin installed with the bluetack blocklist, and had three active torrents, but my settings are pretty modest. It usually takes a few minutes to really start gobbling memory, but I find even 50 MB to be 18 MB too many when I've explicitly set -Xmx32m in /usr/bin/azureus. A java program that ignores jvm parameters makes me wary. That said, if I could keep it at 50 MB, I would probably still use it, but it pretty rapidly gets over that. Since the laptop I'm using only has 196MB, azureus is usually using a lot of swap if I let it run overnight. That's why I call it a memory leak. It seems to be using memory that it shouldn't even have access to, or even a possible use for. Why could it possibly need 100MB? I even followed the azureus FAQ for reducing memory usage -- didn't help at all. It eats and eats.

This is a separate bug and should be filed as such.

On Nov 5, 2007 1:23 PM, Thom Pischke <email address hidden> wrote:

> I do have the SafePeer plugin installed with the bluetack blocklist, and
> had three active torrents, but my settings are pretty modest. It
> usually takes a few minutes to really start gobbling memory, but I find
> even 50 MB to be 18 MB too many when I've explicitly set -Xmx32m in
> /usr/bin/azureus. A java program that ignores jvm parameters makes me
> wary. That said, if I could keep it at 50 MB, I would probably still
> use it, but it pretty rapidly gets over that. Since the laptop I'm
> using only has 196MB, azureus is usually using a lot of swap if I let it
> run overnight. That's why I call it a memory leak. It seems to be
> using memory that it shouldn't even have access to, or even a possible
> use for. Why could it possibly need 100MB? I even followed the azureus
> FAQ for reducing memory usage -- didn't help at all. It eats and eats.
>
> --
> Azureus hangs or crashes showing splash screen at start
> https://bugs.launchpad.net/bugs/57875
> You received this bug notification because you are a direct subscriber
> of the bug.
>

--
Mario Limonciello
<email address hidden>

Of course it is. Just thought it was worth mentioning, since a completely new package is being introduced and tested.

Thom

This is a matter for another bug report, this one is focused on
addressing the crash-on-startup behavior. I cannot readily reproduce the
behavior you describe on a stock Azureus install on my system.

I'm using a custom-installed Azureus (not from apt).

I recently had to remove GCJ completely and a few depending programs I wasn't using anyway because Azureus refused to start.
Azureus was unable to automatically locate Sun Java and found the GCJ one instead.

Canonical cannot support custom installs.

On 11/9/07, Helge <email address hidden> wrote:
>
> I'm using a custom-installed Azureus (not from apt).
>
> I recently had to remove GCJ completely and a few depending programs I
> wasn't using anyway because Azureus refused to start.
> Azureus was unable to automatically locate Sun Java and found the GCJ one
> instead.
>
> --
> Azureus hangs or crashes showing splash screen at start
> https://bugs.launchpad.net/bugs/57875
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>

fre 2007-11-09 klockan 13:03 +0000 skrev Helge:
> I'm using a custom-installed Azureus (not from apt).
>
> I recently had to remove GCJ completely and a few depending programs I wasn't using anyway because Azureus refused to start.
> Azureus was unable to automatically locate Sun Java and found the GCJ one instead.

sudo update-java-alternatives --list
sudo update-java-alternatives --set java-6-sun # <-- insert right java

might work.

>
--
<email address hidden>

Plus ça change, plus c'est la même chose

New deb package works on my Gutsy amd64 pc after I recompiled icedtea-java according to bug #152362:

sudo apt-get build-dep icedtea-java7-plugin
sudo apt-get -b source icedtea-java7-plugin icedtea-java7-bin icedtea-java7-jre icedtea-java7-jdk
sudo dpkg -i icedtea-java7-*deb

Warning: this takes hours! Left .azureus folder in place along with old error files - no crashes, and downloads are no longer crawling. Thanks a lot.

has this backport request being approved?

On Sat, Nov 10, 2007 at 07:28:36PM -0000, Sebastien Bacher wrote:
> has this backport request being approved?
Yeah, it's approved. Please backport it at your convenience,

Thanks, John

I don't really know what happened since I was having the same problems with Azureus but, I just downloaded the latest version of Azureus and installed it over the version that seems to be on the Ubuntu repositories and it works just like it was intended to! For some reason, it seems to be using my Java 5 RE and not Java 6 RE but I'm happy that it's not showing me the start up screen and then hanging when it shows me part of the Azureus display!

Please do not report information that does not pertain to either the Hardy or Gutsy proposed deb posted here. It does not help with testing these packages to state that a package obtained from another source packaged using completely different techniques worked for you. This bug report is already nearing 200 posts and is enough of a mess to tangle through ;-)

Thanks.

For the record, I removed azureus 2.5.0.0repack1-0ubuntu1.1~proposed1 from feisty-proposed. The SRU has never been completed, and I don't see a reason to start caring about Feisty universe now.

 * Trying to backport azureus...
  - <azureus_2.5.0.4.orig.tar.gz: downloading from librarian>
  - <azureus_2.5.0.4-1ubuntu3.diff.gz: downloading from librarian>
  - <azureus_2.5.0.4-1ubuntu3.dsc: downloading from librarian>
I: Extracting azureus_2.5.0.4-1ubuntu3.dsc ... done.
I: Building backport of azureus-2.5.0.4 as 2.5.0.4-1ubuntu3~gutsy1 ... done.

Ok, the 2.5.0.4-1ubuntu3~gutsy1 seems to have propagated to most mirrors, so I'll need everyone's help again for the last step of this update. Please report whether or not this version, found in gutsy-backports, works as well as the candidates you guys so generously tested before. I expect it to be identical, as it came from the same sources, but let's test just to make sure.

After this, it will be a simple copy to gutsy-updates then everyone will have working Azureus and we can move on to more exciting things with our lives.

Cheers,
John

The version 2.5.0.4-1ubuntu3~gutsy1 is working perfectly for me on both sun java and iced tea.

I can also personally verify that the backports built package works as well as my personally pbuildered candidates.