I’ve seen this idea pop up a few times, so here’s my contribution to this meme:
When I was in Outlook, I ran a bunch of mail servers for testing in our interoperability lab. At one point I needed access to an HP Openmail server for some testing, and I found out that the Exchange team had one. So I contacted the owner and asked for an account on the server, which he granted. I quickly found out that it wasn’t configured to allow email to flow outside of the server, so I asked them to do that. He wasn’t sure how to do it offhand and didn’t have time to research, so I offered something like “Well, I have experience with UNIX mail servers, perhaps I could take a look and figure it out?”. They agreed, sent me a huge stack of manuals and gave me admin access, cautioning me to be careful.
Being human, I of course went directly to the part of the manual about “accessing the internet via smtp”, assuming that was the right section. I skimmed through the chapter (again, normal activity here, we all know that no one actually reads documentation ;-) and got to a part where it talked about adding an address routing (or something like that – can’t remember the exact term, this was 6 years ago). The routing address for internet access was something like “*,*,*”. So I added it, tested it, and nope, it didn’t work.
So then I went into the admin program and deleted the address routing that I had just created and was reading through the docs some more when I got a frantic call from a tester on Exchange. Apparently I had deleted the entire server – all of the accounts, mailboxes, etc. Oh boy, egg on my face. But I was also (justifiably IMO!) upset that deleting something I created caused such a catastrophe. To this day I still don’t get how that happened, but it was never a priority to research enough into how openmail works to figure it out.