380517 - Loading Poppit game on Pogo.com causes Camino to crash a few seconds after loading

Status: RESOLVED FIXED

(Plugins Graveyard :: Java (Java Embedding Plugin), defect)

Description

[[e-mail]]

User-Agent:       Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en; rv:1.8.1.4pre) Gecko/20070512 Camino/1.6pre
Build Identifier: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en; rv:1.8.1.4pre) Gecko/20070512 Camino/1.6pre

The game itself loads fine but about 10 seconds after I start playing the entire Application crashes. 

Reproducible: Always

Steps to Reproduce:
1. go to www.pogo.com and click on poppit
2. Click to play the game
3. start playing the game.
Actual Results:  
Camino unexpectedy quits

Expected Results:  
kept running

Comment 1

[Stuart Morgan]

I get a small blank popup titled "Error" and then nothing, so I can't get far enough to try to reproduce this.

Please attach your crash log using the "Add an attachment" like above. Can you reproduce this in Firefox and Safari?

Comment 2

[[e-mail]]

Attachment: The crash report from after Camino Unexpectedly quits.

I'm assuming this is what you were after, haven't actually reported a Mozilla Bug before. As for Firefox / Safari, Firefox works okay, but the animation is slow during the game, Safari has no issues, and this is the only reason I don't use Camino 100% of the time. 

 Let me know if you need anything else.

Comment 3

[Stuart Morgan]

Looks like it's JEP-related, so moving to JEP.

Comment 4

[Steven Michaud [:smichaud] (Retired)]

Yup, this is a JEP bug -- in fact it's a regression that was
introduced in JEP 0.9.6.1.  It (or something like it) happens in all
Mozilla.org browsers on OS X 10.4.X.  But (in my tests) it's only
trunk versions that crash 100% of the time with this bug's URL.

Though it's a crash bug, the problem seems to be very rare:  JEP
0.9.6.2 (which contains this bug) was landed on the trunk on
2007-04-19 and on the branches on 2007-04-22, but I've never seen this
bug before, and this is the first time it has been reported.  There
are no other crash logs in bmo or talkback that contain hitTest in the
top 5 frames, and which therefore might have been triggered by this
bug.

This bug is a side-effect of a workaround (introduced in JEP 0.9.6.1)
for a serious problem in Java 6.0 on all platforms and Java 5.0 on OS
X 10.5 (an applet wasn't garbage-collected if you clicked on it at
least once!).  As best I can tell, it only appears when displayable
objects are manipulated by more than one of an applet's threads at the
same time (I'm talking about the applet's "main" thread and other
thread(s) explicitly created in the applet -- not about threads
created by the JVM, which aren't specific to any one applet).

I've got this problem fixed in my current (working) version of the
JEP.  I hope to release a new JEP version containing this fix sometime
in the next few weeks.

Comment 9

[TommyBee (No E-mail)]

Bug 386349 indicates the same problem at http://www.trussvilleforum.com/ -- click on the chat room link.

Comment 10

[Steven Michaud [:smichaud] (Retired)]

I've just release a new version (0.9.6.3) of the Java Embedding Plugin
that fixes this problem.  Please try it out.

For more information see bug 386918.

Comment 11

[Kevin Horton]

I installed JEP 0.9.6.3 in Camino 1.5 on one computer, then downloaded 12 attachments from different Outlook e-mail messages, with no problems.  This activity would give me a crash perhaps 30% of the time with the JEP that came with Camino 1.5 (see https://bugzilla.mozilla.org/show_bug.cgi?id=383407).

JEP 0.9.6.3 looks to have fixed my problem.  Thanks.

Comment 12

[Steven Michaud [:smichaud] (Retired)]

> JEP 0.9.6.3 looks to have fixed my problem.  Thanks.

You're most welcome.  Thanks for letting me know.

Comment 13

[TommyBee (No E-mail)]

This resolves my issue in bug 386349 as well.

Comment 14

[Steven Michaud [:smichaud] (Retired)]

> This resolves my issue in bug 386349 as well.

Glad to hear it.  Thanks for letting me know.

Comment 16

[Chris Lawson (gone)]

Should this be marked FIXED now? It sounds like, based on comment 11 and comment 13, that this is fixed. The new JEP did get checked in, right?

Comment 17

[Steven Michaud [:smichaud] (Retired)]

You're right, Chris -- the current JEP (version 0.9.6.3) was landed on
the trunk a long time ago, and is bundled with 1.8-branch releases as
of (I think) Firefox 2.0.0.8.  (See bug 386918.)

I'm marking this FIXED.  But it's possible that it should be
WORKSFORME.  If so, someone please change it.