The bug fix was "Don't fecking do it"

Oh boy...

Still it is important to fix the bugs, very important to the customers. I can understand the game needing to meet the deadline and fixing the bugs might comprise a lot of time in other things that still need to be worked on. It still should be fixed.

Was a hardware bug, turned out that the same technique crashed ALL games

He had found a way of dropping the voltage at the 6510 in the keyboard to the point where the chip crashed. This has a chain effect of sending a NMI back to the 68000 and triggering a hard reboot.

Nothing I could have done about it.

I know too many engineers who get something up to about 95% working properly when they're supposed to be finishing tasks and then get tons of praise from management (a) for "completing" work and then (b) fixing piles of bugs later. Of course, not all of them get fixed, and we ship games with these sorts of crashes, while they laugh at stupid QA.

Yes I've seen that too, but I have also seen QA totally melt down.

One guy didn't like the game, so he just didn't test it. He spent an hour a day fiddling with it, then just went on to doing something else. He was supposed to test against all bios versions, he tested against one. Game went out with a huge bug caused by a bios change that had not been applied to my development machine.

The same guy nearly caused me to cover mount a demo on 1 million magazines, with a virus.

He used my machine at night to play pirated games. I didn't take it very well when I found out. I swear I didn't know that wall was only plaster board.

The bug fix was "Don't fecking do it"

Oh boy...

Still it is important to fix the bugs, very important to the customers. I can understand the game needing to meet the deadline and fixing the bugs might comprise a lot of time in other things that still need to be worked on. It still should be fixed.

There are two classes of bugs in that category.

One is "Don't Do That', or DDT bugs. My personal favorite DDT bug was to pause the game, alt-tab out, run the uninstaller, skip the prompts about the game still running, then alt-tab back in the game. It crashes. DDT. Another of my most favorite was when the tester pressed down on the PS2 until the disc ground to a stop during reading, repeating five or ten times until a read error appeared. Thanks for that. DDT. (We joked for weeks asking for permission to grind discs in our very expensive devkits...)

The other is "Known Shippable", or KS. These are always a little troubling, but cannot be helped. These include one-off bugs; the game crashed once doing this, we don't know why and we cannot reproduce it. They also include little annoyances that we can live with; when a player is wearing this clothing combination, is running and turning left, some of their clothing polygons clip through each other.

As the deadline approaches, the number of KS bugs rapidly increases. In the days right before submitting to certification almost every new bug goes straight to KS status; if they haven't reported it with three months of testing and it isn't a crash bug, we can probably live with it.

I hate crash once bugs. You spend ages trying to recreate them, add loads of debug code to try and track it down and you never see it again.

There has to be a reason.

When I worked at Panasonic we had one of those, we created a special build with loads of trace information in it. Then everyone in the company took a handful of phones and went somewhere. Like the true coder I am, I went to the pub. Ever 3 minutes we dialled the speaking clock on all handsets. We made so many calls, Vodafone cut us off.

Shut down all the companies phones.

Eventually someone got a crash, actually at their home. A team with loads of test kit jumped in a van and camped in his front room until the problem was found.

That's the sort of massive effort required to find bugs sometimes.