@glennf those are usually fine for relatively simple things like converting or querying data, especially for single users (see also: Excel) but all of them quickly become unwieldy or unusable when there’s any sort of real complexity
@thomasfuchs It's weird, because there is something like 30–40 years of visual programming systems that can be pretty robust, or at least address many people's need for data handling. Those are typically vastly better than all the "solutions" even when they're imperfect.
@thomasfuchs That part of “The Psychology of Computing Programming” might be read that way now but it’s not what Weinberg was talking about in context. He was referring to the emergence of software methodologies that didn’t take into account psychology.
@MartyFouts I'm looking forward to the rest of it, this just really jumped out at me–a thinly veiled diss of executives for not understanding what they're doing.
I've read other Weinberg books before but not this one, enjoying it so far.
@thomasfuchs oh AI researchers have always over promised but under delivered, starting with Herbert Simon’s 1956 claim that human translation would be eliminated in 10 years. Surprisingly few books have addressed this. But Weinberg was concerned with something far more fundamental: the failure to treat programming as a psychological activity. That failure continues to this day.
@MartyFouts The thing is, humans don't change. I'm sure it could be adapted relatively easily, and yeah I guess PL/I isn't the greatest language for examples…
@thomasfuchs Its examples are dated but they were well considered for the time. I have long thought it needs an update to accommodate how the sociology of programming has changed. I am sure that you will like it.
@tony In their defense, I think what some of these things did is lower the barrier of entry and made it easier to learn how to program; even if that wasn't the intention.
AI on the other hand—I think it will make learning worse because to learn programming you actually need to sit down and program and fail and iterate etc., because that's how brains work.
@thomasfuchs Unfortunately, computer executives will be computer executives.
As reported by the The Verge, there are currently layoffs going around in the industry. There could be many explanations, I could however fear that one explanation could be that a lot of executives thinks AI will mean they can do the same work with a lot less employees. They will be mistaken, if that is the case, but it might be a painful period until they realize that.
@thomasfuchs I am not sure if that is better (that the executives know) or worse (that they know and do it anyway).
As for the recent wave of layoffs, I am once again amazed how the supposedly highly innovative and trailblazing Sillicon Valley keeps showing amazing flock mentality.
@thomasfuchs —while y’all digging through that box in the attic, does anyone remember the CASE (Computer Aided Software Engineering) promised iteratively through the `80s & `90s? :ablobgrin: