My favourite one was when I asked the user “can this happen?” (Some value being negative) and they reply “no never”.
Then, of course, I get an occurrence the day of the live demo with the user and her boss.
I ask again, “uh, so is this normal? Has it ever occured before? Because I asked you if it could happen and you said never.”
Now the boss replies “oh, we meant it’s extremely uncommon. Almost never happens”.
Turns out it happens once every few months, amongst hundreds and hundreds of transactions.
So I gently explained that the computer doesn’t care how often it happens. If it can happen, I need to code it in otherwise things go wrong!
Thankfully I had planned the eventuality, so I had a nice error message, but still. A lesson was learnt that day.
It pisses me off when I have to explain to lead developers that we do in fact have to care about things that are “unlikely” to happen because they can in fact happen and when they do they will cause downtime just because you didn’t want to spend 16 man hours adding a solution to the problem
My favourite one was when I asked the user “can this happen?” (Some value being negative) and they reply “no never”.
Then, of course, I get an occurrence the day of the live demo with the user and her boss.
I ask again, “uh, so is this normal? Has it ever occured before? Because I asked you if it could happen and you said never.”
Now the boss replies “oh, we meant it’s extremely uncommon. Almost never happens”.
Turns out it happens once every few months, amongst hundreds and hundreds of transactions.
So I gently explained that the computer doesn’t care how often it happens. If it can happen, I need to code it in otherwise things go wrong!
Thankfully I had planned the eventuality, so I had a nice error message, but still. A lesson was learnt that day.
It pisses me off when I have to explain to lead developers that we do in fact have to care about things that are “unlikely” to happen because they can in fact happen and when they do they will cause downtime just because you didn’t want to spend 16 man hours adding a solution to the problem