starting a thread twice

[Marlene Miller]

class A implements Runnable {
  public void run() {
    System.out.println("been here");
  }
}
class Test {
  public static void main(String[] args) {
    A a = new A();
    Thread t = new Thread(a);
    t.start();
    Thread.yield();
    System.out.println(t.isAlive());
    t.start();
    System.out.println(t.isAlive());
  }
}

Output:
been here
false
true
I do not understand why an IllegalThreadStateException does not occur at the second t.start().
Even so, I do not understand why the second t.isAlive() is true.

[Jim Yingst]

Looks like this has been a bug in the JDK's since, well, forever. Last milennium, anyway. See 4773384 for the most up-to-date report I could find. They say it will finally be fixed in 1.5 - but I'll believe it when I see it. :roll:

[Marlene Miller]

Thank you Jim for doing the research I should have done. I forgot about the bug database.
They say the JVM treats the second start request as a no-op. It's strange isAlive becomes true.
[ June 30, 2003: Message edited by: Marlene Miller ]

[Jim Yingst]

Yeah, back when I sterted Java with 1.1, about a third of my questions were best answered by checking the bug database. Fortunately nowadays that happens much less frequently; it's easy to forget to check...