This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.

Bug 130383 - AssertionError at
Summary: AssertionError at
Alias: None
Product: guibuilder
Classification: Unclassified
Component: Code (show other bugs)
Version: 6.x
Hardware: All All
: P3 blocker (vote)
Assignee: issues@guibuilder
Keywords: RANDOM
: 180298 (view as bug list)
Depends on:
Reported: 2008-03-18 11:51 UTC by Petr Dvorak
Modified: 2010-02-25 08:21 UTC (History)
3 users (show)

See Also:
Issue Type: DEFECT
Exception Reporter: 37796

stacktrace (3.21 KB, text/plain)
2008-03-18 11:51 UTC, Petr Dvorak

Note You need to log in before you can comment on or make changes to this bug.
Description Petr Dvorak 2008-03-18 11:51:45 UTC
Build: NetBeans IDE Dev (Build 200803170003)
VM: Java HotSpot(TM) Client VM, 10.0-b19
OS: Linux, 2.6.24-12-generic, i386

User Comments: 
Simple: More encapsulated JPanels. More elaborate: I created a new Java Application, and I insertd new JFrame form. Then, I inserted a JPanel in the form (and I manually docked it in the bottom) and I changed the color of this panel to Green. After that, I inserted one more pannel, changed its color to red and I docked it to top of the form. I inserted two JButtons and one panel (and changed its color to yellow) in the upper (red) panel. When I attempted to insert a new panel in the yellow panel (which was contained in the red panel), exception was thrown.
Comment 1 Petr Dvorak 2008-03-18 11:51:49 UTC
Created attachment 58564 [details]
Comment 2 Jan Stola 2008-03-18 14:35:16 UTC
I am sorry, I am not able to reproduce this issue. Could you, please, provide more details. It would be ideal to attach 
the corresponding .java and .form files (saved in the last good state) and describe _exact_ steps that should be done 
to see the exception. Thank you in advance.
Comment 3 Petr Dvorak 2008-03-18 15:45:37 UTC
It is not reproducible after regular (I didn't find scenario) steps. However, I did reproduce it. Just put like 5 panels
in the JFrame and play with them: resize, remove, change their order, encapsulate them to each other, ... just play with
the panels in the JFrame...

Adding a keyword RANDOM
Comment 4 Jan Stola 2008-03-18 17:47:20 UTC
I doubt there is anything random in this problem. The layout algorithm used in GUI builder produce always the same 
result, but the result may depend heavily on small details (e.g. where exactly is the component dropped etc.)

It doesn't help at all to suggest me "just play with the panels in the JFrame...". This bug report is useless without 
exact steps to reproduce.
Comment 5 Petr Dvorak 2008-03-18 18:19:36 UTC
Well, first - I adore your attitude. "The bug report is useless" - it was submitted using built-in bug reporter, demand
your colleagues.

Second, you made a great observation - computers are in principle deterministic machines, so anything that happens there
is potentially reproducible and there are definitely no "random" issues (if we don't count the cosmic rays and stuff,
that are deterministic as well though).

However, if you follow the guidelines from, you will see that this
issue matches the definition of RANDOM, as it speaks only about ability to reproduce and frequency (that's what random

When anyone is able to reproduce the issue, he/she will remove the RANDOM keyword, up till now it remains there.
Assigning a RANDOM keyword again. If you think the issue is so easy, you will be definitely able to reproduce it from
the stack-trace, just like anyone else does.

Second, when I work with the IDE, there is like 20 exceptions a day. If I was just a regular user of NB, I would be
pretty frustrated. The suggestion was quite accurate - you can get an exception quite often (I don't even send all of
them) if you just play with the panels.

Finally, I will do my best to reproduce it, as soon as I can manage to do that, I will let you know.
Comment 6 Jan Stola 2008-03-19 10:23:18 UTC
OK, let's clarify my last comment that made you so offended.

I don't want to start a flame war or philosophical discussion about randomness in computer science. I believe that 
RANDOM keyword belongs to issues that are not always reproducible if you repeat the same user gestures in the IDE e.g. 
the result may depend for example on thread scheduling (e.g. things that user cannot affect). By removal of the keyword 
I just wanted to point out that this is not the case here. I see that you have a different oppinion on that. Hence, 
feel free to keep the keyword there if you want to (it doesn't harm anyone). The essence of this comment was to make 
you aware about subtle differences when placing/resizing/moving components in Free Design. If you concentrate on them 
then you have a better chances to provide reproducible test cases.

> If you think the issue is so easy, you will be definitely able to reproduce it from the stack-trace,
> just like anyone else does.

Some problems can be identified just from the stack trace and make sure that I always scan the source code to guess 
what can cause the problem. Unfortunately, it is not possible in GUI builder quite often. The exception is usually 
thrown by some low-level algorithm that got into an unexpected state and it is not clear (without the knowledge of the 
context e.g. the form) why this occured and what should be done in this case.

This is also the reason why this bug report is not usefull without exact steps to reproduce, unfortunately. Note, that 
it does not help to tell me to find the problem myself. Why would I need a bug tracking system and quality engineering 
with this approach?
Comment 7 Petr Dvorak 2008-03-19 10:41:57 UTC
jstola, you don't need to worry :o) - I was not offended. I was more surprised - the stack trace tells you that it is
not an exception, but an assertion error. Furthermore, there are modules and exact line numbers specified. I didn't
expect any complications with resolving this issue then.


I submitted the issue using the built-in exception reporter, I could have also just let it go, but " Why would we need a
bug tracking system and quality engineering with this approach?".

I got the same issue about 3-4 times just yesterday - it appeared intermittently and I didn't manage to reproduce it.

> RANDOM: This issue occurs intermittently and cannot be reproduced reliably.

If you are at work we can discuss it on coffee break.
Comment 8 Jan Stola 2008-03-28 15:47:36 UTC
Closing the issue as non-reproducible for now. Feel free to reopen if you are still able to reproduce it, but don't 
forget to provide exact steps to reproduce then. Thank you in advance.
Comment 9 Petr Dvorak 2008-06-24 12:44:02 UTC
Comment 10 Exceptions Reporter 2008-10-21 08:28:55 UTC
Reopening - reproduced in NetBeans IDE 6.5 Beta (Build 200808111757)
Comment 11 pribyl 2008-12-05 17:30:47 UTC
This issue was reopened by Exception reporter, unfortunately no additional info that could help to reproduce the problem
can be found in the report.

I still cannot reproduce this :/

Closing for now as WORKSFORME.

If anyone manage to reproduce this reliably, please reopen. Thanks
Comment 12 Petr Dvorak 2009-01-08 16:37:17 UTC
Comment 13 Jan Stola 2010-02-25 06:51:53 UTC
*** Bug 180298 has been marked as a duplicate of this bug. ***
Comment 14 Jesse Glick 2010-02-25 08:21:28 UTC
(In reply to comment #11)
> This issue was reopened by Exception reporter, unfortunately no additional info
> that could help to reproduce the problem
> can be found in the report.

That just means that the assert statement is inadequate. The form module needs to be fixed to somehow audit actions that are taken by the user and their effect, so that if and when a seemingly impossible situation is encountered, the assertion error includes all the information needed for a developer to diagnose the problem.