The prank cursor that resulted in an employee being fired before they even started
Raymond Chen
I learned the story of someone who got fired even before they started. It resulted from a prank gone awry.
The perpetrator in question was completing an internship and committed code into the Windows 3.1 code base that was a little prank for the test team: Under a very specific error condition, it changed the index finger pointer to a middle finger.
This was a very rude prank, but the test team would take it in the spirit it was intended.
He then ended his internship and took a two-week break before returning as a full-time employee.
It so happens that the build in question was selected to be a beta release that went out to hundreds of hardware and software companies, as well as beta testers. It wasn’t long before some of these copies found their way into the hands of the media, and the rogue cursor reportedly nearly made it into a screen shot in an industry magazine.
As you can imagine, this created quite an uproar, drawing the involvement of senior management. Executives from partner companies were furious that such an unprofessional image could be included in the product, and I’m sure a good number assumed that Microsoft intended the rude gesture to be directed at them specifically.
This was a scandal of the highest proportion, and someone must pay.
The manager of this component replied, completely truthfully, “the individual responsible for this regrettable act is no longer with the company.”
This helped calm the storm.
The person who made the unauthorized change rejoined the company a few weeks later, as originally planned. I suspect there was a very stern talk as part of the onboarding process.
Bonus chatter: A bug was filed in the RAID database to track the problem and its resolution. In the bug, there was some discussion as to how the issue should be classified. Was it an “off-by-one” error? Or maybe it was a “bad pointer”.
Light
Dark
8 comments
I got caught by something like that once. I’ll leave the exact details under wraps but I was asked to “refrain from putting Joseph Stalin quotes in the user manuals”.
Yup, definately a case of “bad pointer leaked into the public”.
And IMO, this case is caused by arbitrary value. And bug caused by arbitrary value, even if just off-by-one from the value it should be, shouldn’t be classified as “off-by-one” error.
I think you missed the humor. There is a cursor with the pointing finger. This one uses the next finger. Therefore, it’s an “off-by-one” error.
No, I didn’t miss it.
The correct value is cursor with pointer finger, but he set it with cursor with middle finger. That’s what we call arbitrary value.
Say, if a function should return “Message ABCD” that is internal hash value of 1234, but I returned “Message FGHI” that has internal hash value of 1235, it’s still not classified as “off-by-one” bug.
It’s a joke.
That is a failure of the QA and code reviews.
If the intern can slip through a prank, good for them. This actually helps strengthen the company processes.
I can understand the “temporary firing”, but hope Microsoft actually held onto such an individual for the long run.
I think the time of Win 3.1 “code review” was not even exist as established software development concept, and it’s unclear whether the build was selected as beta release before the test team have chance to fully checked it (we know beta software are unfinished ones and can contain bugs, so a bit unrealistic to require test team check it fully before release it. Worse, the test team head shouldn’t have known about the prank, and possibly said it’s okay to release when upper management asked about it, based on prior progress of testing).
I’ve certainly known builds selected for beta testing by a customer that didn’t even install properly.