I think it's inevitable if the early users are incentivized to have their early expertise via means of their own economic survival. At the same time this also makes sense, because there's always a risk in trying something new, there also needs to be a reward.

SQL actually is used by non-developers. When I search indeed (SF bay), I'm seeing a pretty broad range of roles, including one in Customer Support on the first page of results

A point to note in that job post is: basic SQL queries. So, some SQL is used by non-developers

I've come to believe that any system which requires mental simulation of a complex process that can't be seen will always remain exclusive to "experts". Textual coding (including SQL) requires this in at least two ways: mentally simulating a parser/lexer, and then mentally simulating the process described. Mentally simulating something complex and invisible (playing chess, coding, math, chemistry, counting cards, ...) probably can't ever be done without extensive training. I think because of that, it will only be done by those with extremely strong need for it (i.e. high enough to justify hundreds of hours acquiring a new skill), or who are lucky enough to find the effort of training rewarding in itself instead of a cost. Also it's a core thesis of mine that simply making the invisible, mentally-simulated stuff visible could massively lower the barrier and admit many more people into the field!

That quote about SQL is from a reunion in 1995 reminiscing about what… the early 70s? Given the user interfaces and interactions at the time, I’d say they did a pretty good job!

I know people with multiple degrees who struggle to connect HDMI cables, a task barely more difficult than a square-peg-square-hole baby toy. When it comes to getting started, it's not about how difficult the task is, but the narrative the person has about the task. You can tap and click stuff today, so people's narratives about typing stuff has changed. At the time SQL was invented, a non-programmer would be what—someone who could use DOS but not Fortran or BASIC? If Phyllis Reisner ran the same human-factors tests today, SQL would probably appear to be less usable than it was in the '70s. Actually, that'd be a fun study. 😆 Out of curiosity, I browsed a few beginner SQL tutorials just now. Five didn't say anything how easy it is to get started with SQL, or imply in any way that SQL was designed with human factors in mind. The sixth does point out that, hey, there's a million database software packages out there now and most have custom commands, but all you gotta learn is six commands (with a list of their simple, evocative names) and you'll know pretty much everything you need to know about SQL. Now that's a tutorial for non-programmers about a language designed for non-programmers.

@alltom i would love to read that last tutorial if you still have the link

Now that I've read past the intro, I wouldn't exactly endorse this tutorial, but here it is. :)

ahh dang i really liked the framing device

Hmm, anyone else here always forget the syntax of INSERT and UPDATE because they are so different from each other? Doesn't seem like great attention to human factors in that case.

Yes! I’d be curious to know why they are different.