Garbage, eat your own garbage!

see this link: http://www.itjungle.com/fhg/fhg051204-story05.html

and this text in it:

GOOFY PTF'S

When I read through IBM's various PTFs, sometimes I have to laugh. I can just imagine the stories behind some of them. Take, for example, PTF MF32064 for Licensed Program 5722999 for V5R2M0. An application server ended abnormally during interaction between Java interrupts and garbage collection. This didn't sound too strange. Any programmer worth his salt knows that when you make a change in one program, the chances of causing problems in another program are pretty good. And with the complexity of the programs that make up OS/400, those chances increase exponentially. But what was funny to me was the rest of the problem description: "It takes an unusual situation to cause the problem: there has to be a thread which was interrupted for garbage collection during a wait or sleep, went into another wait while processing the garbage collection interrupt, and then was interrupted yet again."

You can just imagine the hell that IBM tech support and programmers must have gone through in verifying that this problem even existed and then trying to recreate it to make sure that the fixes they had made did in fact fix the problem.

IBM Supervisor: "What are you working on?"

IBM Programmer: "Writing a program that will interrupt a thread, but only during the garbage collection process. Oh, yeah, and the thread has to be in a wait state. Oh, and then I've gotta figure out how to cause another garbage collection cycle, but only after the thread goes into another interrupt."

IBM Supervisor: "How do you even know the customer's description of the problem is accurate and that the program you are writing will duplicate the problem?"

IBM Programmer: "I don't. I first have to write a program to see if the problem even exists before I can write a program to fix the problem. Assuming it exists."

IBM Supervisor: "So…I should call your wife and tell her not to wait up for you?"

Sometimes I'm really thankful that I don't work in software support. For more information on this PTF or the Java Group PTF it belongs to, go to IBM's Web site.