Unacknowledged Java bugs

[Larry Barowski]

Does anyone know of a site that lists Java bugs that are not
acknowledged by Sun? I did a bit of searching but couldn't
find anything. I'm particularly annoyed with their response
to:
http://developer.java.sun.com/developer/bugParade/bugs/4773491.html

I resubmitted this and got the usual "we have determined that
this is a new bug" response, but it never showed up. The bug
(IllegalComponentStateException thrown when clicking
on various components in a dialog while it is closing) may be
harmless for all Swing components, but I would guess it does
cause a problem somewhere. Even if it does not, it could in a
subclass. I may be paranoid, but I suspect the reason this bug
is ignored is because it rarely occurs but fixing it would
require a lot of changes to the code.

Also useful in such a list would be bugs that Sun considers
"not submittable" without a reproducible test case, but for
which generating such a test case is impossible because
a) it is a stack dump sent by an automatic reporting system,
and not reproducible by the application developers, or
b) when a small test case is written for a bug found in a large
application, the bug mysteriously disappears. In many cases
it is fairly clear from a stack dump that the bug is in the Java
libraries and not the application.

If such a site does not exist, we'll consider hosting one.

 

[rpnman]

To answer your question, no, I do not know of any such web site.

But I have a more general observation to make, and a question or two.

I can understand your frustration, but, in fairness, I have to observe that
"State: Closed, will not be fixed" sounds to me like an acknowledgement. If
your "resubmission" of the same bug was ignored that is an entirely
different matter. As a developer, I put pretty high thresholds on the
re-opening of bugs.

In the year and a half since you submitted and it was closed, two others
have posted to identify the same thing. Not exactly overwhelming demand, but
some interest.

I guess I also think it should be reconsidered also, at this point. The
original response would probably be better if more status were attached. The
evaluation says: "Not common usage. Not a feature. This seesm [sic] to be
timing related." This assessment may be accurate, but seems too brief and
dismissive for my tastes. It also seems to me that the code should handle
that kind of abuse by users.

However, I notice there is no way to vote for a bug to be re-opened. Perhaps
there is no provision for that to be done at all. (As a developer, I can
understand not wanting to re-open bugs because 2-3 people don't like the
disposition, but...)

Finally, my questions: When you "resubmitted" it, did you reference the
original post? When did you re-post?

 

[Larry Barowski]

To answer your question, no, I do not know of any such web site.

But I have a more general observation to make, and a question or two.

I can understand your frustration, but, in fairness, I have to observe that
"State: Closed, will not be fixed" sounds to me like an acknowledgement. If
your "resubmission" of the same bug was ignored that is an entirely
different matter. As a developer, I put pretty high thresholds on the
re-opening of bugs.

"Not common usage. Not a feature" makes me think it is not an
acknowledgement, or a least a misunderstanding. This exception
can happen when closing almost any dialog in any application, if the
user intentionally or accidentally clicks on the dialog as it is closing.

In the year and a half since you submitted and it was closed, two others
have posted to identify the same thing. Not exactly overwhelming demand, but
some interest.

Demand for a fix in a library should not be a consideration when
something as basic as this is broken and there is no reasonable
work-around.

...
Finally, my questions: When you "resubmitted" it, did you reference the
original post? When did you re-post?

Yes, I referenced the original. That was months ago.

At least that bug is on the site and developers can find it. Since
Sun currently does not accept most bugs without a test case
(and I can understand that), it would be good to have a site
listing stack dumps that developers believe to be caused by
bugs in the Java libraries, but can't reproduce. That would allow
the pooling of information to help determine the cause and what
to do in the case of the problem exception. Since our app. has
automatic crash reporting we often get stack dumps and the users
tell us what they did and whether or not things seemed to work
after the exception. Stack dumps from netbeans are copied to the
Java bug list without test cases, and three or four times that
information has saved me effort by at least letting me know it is
probably not my problem. For example:
http://developer.java.sun.com/developer/bugParade/bugs/5027059.html
is one that's been hitting us, but I have not been able to
reproduce it.