Troubleshooting as a core testing ability: a true story


An ability all testers need is troubleshooting.  When presented with  a problem, testers need to be able to find the cause to the narrowest possible instance.  Finding steps to reproduce a bug is critical - it does no one any good to say "Application foo crashes every so often."  A good tester will narrow down the conditions that are occuring to get to the root of the problem.  It can be tedious (very tedious), time consuming, frustrating and complicated in ways you never imagined at the onset of the investigation.  On the other hand, it can be very simple: so simple, you overlook the problem completely.  This is one of those cases.


Last time, I mentioned I started in technical support for Windows 95.  Ah, the joys of dealing with serial devices and fighting for interrupt lines.  A fellow calls one day and says, "I just got a new mouse, and it conflicted with my modem.  So we pulled the modem out, rejumpered it to use IRQ4, set the mouse to IRQ3 and now the mouse works but not the modem."


I set out with the standard questions first: internal or external modem (internal), what brand (Cardinal), had it ever worked (yes), and whether the phone line we were using was routed through the modem (it was).  "Great," I thought, "there's a known problem with Cardinal brand modems using IRQ4.  Let's investigate."  So we do.  We "investigate" for 20 minutes or so.


(And astute readers will already have seen the problem and are yelling the fix to me.  Sigh.)


We get everything set back up again and try some DOS echo commands to see if the modem makes clicking sounds on the phone.  Nope.  We keep after it, trying the small matrix of interrupts being used by the two devices, and finally decide to unplug the mouse to see if that helps.  It doesn't.


So I tell the fellow to re-jumper the modem to use IRQ4.  He immediately says "OK, it's set."  I think this is  too fast since I know how long it takes to move the jumpers around.  He says this modem actually has little switches on the back and moving them to IRQ4 is very simple.  He even uses a wood toothpick to move the jumpers since he fears static  electricity so much.  He even has the modem sitting on a static free mat while we troubleshoot….



Yes, he had his internal modem sitting on a mat external to the computer. He even said this first thing: "So we pulled the modem out..." and never mentioned "we plugged it back in."



I pointed this out to him and you could hear the proverbial pin drop.  He simply said, "Oh, yeah, I guess I need to plug it back into the slot in the computer."  He had to hang up to do that.  I assume it worked since he never called back.  And of course I was upset with myself for spending 45 minutes of our time while overlooking the actual problem.


Everyone should have to do tech support to work in the computer industry.


Comments, questions, concerns and (this time I'll probably get a lot of) criticism,



Skip to main content