I do know better, honest.

I knew I shouldn't have done it. I knew I should have tested the computer lab worksheet on the machines in the computer labs, rather than just testing on my own machine.

Result? Half an hour's worth of firefighting in the lab session today, caused by:

• 1 missing file containing library code that needed to be tracked down and imported specially
• 1 library procedure that had gone missing entirely (fortunately there was a convenient substitute to hand)
• 1 typo in the sample code that I gave to students, so that it wouldn't compile

I do know better, really. With a completely new worksheet, I know to

1. do the worksheet myself before I set it,
2. do the worksheet again myself from scratch on my machine,
3. publicise the final draft of the worksheet in advance to any of my colleagues who are going to have to read it anyway (so their preparation might as well be combined with telling me about worksheet typos before the lab session, rather than afterwards), and
4. take the worksheet down to the lab (ideally some days later) and test it there all over again.

This method is pretty good at ensuring that typos are down to a minimum, there are no major surprises waiting for you in the labs that could have been predicted in advance, and any remaining errors are usually some combination of {small, ignorable, fixable, only affecting a few students}.

For this worksheet, it took way longer to develop than I thought it would, I ran out of time and so couldn't do parts c) and d), and I paid the price. Oh well. It was still mostly a good worksheet, the students seemed to get a lot out of it, and it was time well invested, given that this worksheet can be used in subsequent years.