I'm filing this bug report in response to correspondence with a user
who has been working with us on some unrelated issues with Ladybird.

I'm not sure of the cause of the crash. We have had a couple of reports
of crashes on startup from CAP members, so I am giving this a high priority.
The fact that he saw occasional crashes with Blackdown's 1.3 is not
necessarily reason to disregard this report; the crash he reports below
was seen with Sun's JDK1.3.1-beta.

Here is the correspondence (see Comments for entire email with headers):

============================================================================

But now, I gave the Beta a test. As before, I have done that on the same
machine as last year, when I sent the bug-report (PIII, 550 MHz, 128 MB),
and OS (SuSE Linux 6.3; Kernel 2.2.13), AND A GERMAN KEYBOARD.

I have installed the Beta and set JAVA_HOME accordingly. Since last year
the problem showed up when I tried to use forte4j, I have used that test
again. As a result, I can tell you that most, but not all of the problems
seem to be fixed. But let me tell you step by step:

* I called the forty4j start-up script, which is appended as attachment 1
* at first, the start-up seemd to be OK, but then, at some time during
  loading of modules, the app crashed; it produced a lot of output on
  stdout/stderr; I append the important parts of that output as attachment
  2, with some comments interlaced (<<< ... >>>); I hope, you can get
  the essentials
* without change, I tried to start forte4j again, this time re-directing
  stdout to a file; I append that file as attachment 3; this time, f4j
  started correctly, but with two warnings on stderr; I append those
  together with an info message on stderr as attachment 4
* within f4j, I seemingly could enter ALL of the special chars, which
  I couldn't last year, e.g., {}[]~@
* executing the test program with those chars entered to a string,
  yielded again one error message, together with the expected output
  string; I append the source-code of that test program as attachment 5,
  and the program's output as attachment 6

I hope, that report will help you, to track down the remaining problem
with respect to my bug-report.

In my oppinion, the crash during the first start-up was some problem with
thread synchronization on my linux machine, but I'm not sure, what caused
that problem. Every once and again, I have such a crash with the blackdown
1.3 version, I use since last year. But that are rearly very sparse
occasions, in the order of three or four all in all during 6 month or so
of daily work with the jdk. Very good work, I must say. Work done by you
and the guys from blackdown!

Filed by ###@###.### on March 24, 2001.