Saved by a fax machine: the most ridiculous error I ever made with a computer

tl;dr version: I stuffed up a computer by mucking about with the regedit or .bat file and it wouldn't start. This was in the early 90s and the only way the company could help was by faxing me instructions to type into a new text file to save on a floppy disk from which I could then boot up. 
Fortunately it worked :)

In the early 1990s I used a computer system which controlled a chromatography pump*, for science.

At some point something went a bit wrong with the system and my boss suggested that I be a bit braver than I had been about fixing it myself so I read the manual and asked people in the computer department. I learned that I had to do something to the registry file, which underpinned the whole functioning. So I did.

After I'd done what I thought I was supposed to do the computer wouldn't switch on (well it wouldn't boot up and I couldn't interact with it), so I was now in a worse position. By now my boss agreed with me that I probably should have called in an expert and I was a bit worried that I'd seriously stuffed up the computer and rang the manufacturer to ask for help.

The company said that I'd need to boot the computer from a disk (which I didn't have) so they said they'd fax me a set of instructions - I don't think they had email at that time, though I'm fairly sure that I did (was working in a university), so a fax it was. The fax turned up and the program was pretty short - I went to another computer, opened up a .txt file in notepad, typed in the code and saved it with the appropriate file ending onto a floppy disk, put it in the moribund computer and switched it on. It worked perfectly ;)

I am just recording this small curiosity in the history of me killing computers...

*The pump gently delivered a stream of solvent, at a defined rate, through a long thin chromatography column which I used to separate components in my samples, for lipid chemistry purposes. The column contained a substance that slowed down - at different rates - all of the components in my samples as they passed through. This meant they came out the other end ('eluted') at different times and the amount of them could be measured individually. The separation was based on a relative attraction to either the solvent or the column's retarding material (also a little bit based on their size and other physico-chemical properties). This resulted in a complex mixture going in one end and individual components 'eluting' (it's a good word!) from the other end, for me to collect and see 'how much'. The computer provided a reading of the output based on the refractive index of the eluted solution (eluent), transferring this to an on-screen graph.