There are spreadsheet authors who would have been programmers in "another life" and it's usually possible to get thru to them re: the kinds of issues that plague spreadsheets. Those people are a pleasure to work with. I remember showing a junior finance guy how to use "JOIN" to overcome performance problems with his ad-hoc VLOOKUP-based re-implementation. He "got it" and things got a lot better.

OTOH, there are people who have no background in computational thinking or software development churning out spreadsheet nightmares w/ no regard to accuracy or maintainability. Anecdotally, they seem to be the people most attracted to glitzy formatting features, "no code" automation tools, and frightening nightmares of "linked" spreadsheets.

▲

f30e3dfed1c9 2 days ago | parent [-]

"OTOH, there are people who have no background in computational thinking or software development churning out spreadsheet nightmares w/ no regard to accuracy or maintainability."

FWIW, I think this should be seen as a failure of the software industry more than as a failure of those people. Pretty obviously, the vast majority of people have "no background in computational thinking or software development" and so they use what they can and in many cases, that's a spreadsheet.

If the software industry saw fit to offer something that was technically better while still being something they could actually use, they'd use it. But it hasn't.

	▲	EvanAnderson 2 days ago \| parent [-]
		I certainly didn't mean to sound over pejorative about the "non-programmer" spreadsheet authors. They're people trying to get stuff done using IT automation and that's a laudable goal to me. Sometimes their best intentions end up making nasty work for others later. Their hearts are in the right place. I have mixed feeling about spreadsheet nightmares as a failure of the software industry. I don't think any amount of well-designed, technically better software is going to attract the particular stripe of people who see the ends w/o concern for the means. I feel like that kind of person is usually unwilling to learn anything that doesn't appear immediately in service of their goal (even if that learning would end up helping them in the future). Short of AGI that to dictate their requirements to I don't think there's a technically better solution for some people. Even then, AGI would have to ask clarifying questions which, in my experience, this type of person regards as tiresome and needless. Trying to get these people to discuss corner cases, hard-and-fast business logic, and even basic workflow is painful.