The new programmer, Mark, was straight out of university but soon found his feet. The nice suit of the first few weeks had been replaced by t-shirt and leather jacket with a custom car design on the back. He was contributing a lot of code - and opinions.
The project seemed to be going well, lots of code written and it all seemed to hang together. There was a good team spirit which meant that after hours we'd shoot the crap out of each other playing Doom.
This then lead onto us trying to shoot the crap out of each others code. Mark made my ears go red as he loudly announced to the room the crash he'd found in the area that I was working on.
Before I tried to fix it I had to get some payback so I tried HIS code out and found a crash
Take that !
Whilst I was trying to fix my bug I heard Mark triumphantly proclaiming that he'd fixed his and why was I being so slow
No-one likes a smartass so tested his fix and found another bug
And then another
And another
Soon he was buried under a deluge of bug reports
Which got me thinking. Why wasn't our tester finding these bugs ? Everyone seemed to think that the project was in good shape - they were even busy designing a logo and packaging and making plans for how to handle support calls
But I was crashing the program at will and in no longer than 5 minutes.
So my coding efforts started to diminish and I was spending more and more time testing the program, trying to reproduce and narrow down bugs and generally hammering the system. The supposed tester got the boot - the fact that he was bringing porn into work and spending his time looking at jugs rather than bugs didn't help his case.
As I was a programmer on the project and had access to the code and knew how it hung together I was able to change the code and soon I had some primitive automation going and last thing at night I would set 6 machines off and running and was often awake before the alarm so I could hurry into work and see how many had survived the night.
The outcome of all this ?
A 4 mice( out of 5 ) review from a Mac magazine that praised the programs stability ( this was back in the days when Mac users expected programs to crash )
The best annual review I'd had at the company - there it was in writing "Phil's testing efforts saved the project from disaster"
Sadly the program was a commercial flop and it was time to start a new project
And back to coding
