I messed around with that in the 1990s but no longer remember anything.

I think I had only one book, "The Professional Programmers Guide to Prolog" by Alan Hamilton. I still have some of the programs from it somewhere.

ETA: I also had "Prolog Programming" (2nd Ed.) by Ivan Bratko

Thanks old bean.

Was it 1.0 or 2.0.

Your forgetfulness isn't helping me. Do you take supplements 🤣 ?

Was just checking that now. It was 1.0 or 1.1 - I don't recognize the interface of 2.0.

So you can write stand alone programs with it?

Everything I've heard says you present a database if you will, then run a query over it. Ala 80s type expert systems (which died a horrible death it seems).

Yes the interface, which suggests it works something like an interpreter. But maybe that's just a pre complile environment? Bear with me as by now it's obvious I'm totally clueless.

Yes, just an interpreter. I don't remember executables.

ETA: yes you can create executables, but I don't believe I ever did. I only ran programs using the interpreter.

I played with Win-Prolog or LPA-Prolog now and again. Not for quite a long time now. It seemed a clumsy language when trying to write the sort of code I might have written in BASIC or Pascal - Python is much more natural by comparison. As I recall Prolog revolves around a database of rules that it iterates through, and you kind-of ask it questions and then it chunters through it's rule set, and you have various ways to stop it iterating at certain points. So yeah, it was couched a something for AI and databases. But for those of us dabbling it can be a weird, mind-boggling language to use for the sorts of tasks we might normally write code for. I tried to write a sudoku solver in it, as that seemed the sort of thing it might be useful for, and I was into sudoku at the time. It took me weeks on and off and in the end I gave up - my code would get so far correctly, then do a kind of prolog brainfart in some way I can't remember (like it iterated twice and got confused, or couldn't get beyond a certain point). I then rewrote my algorithm into Python (which I had the simplest beginner's understanding of) and the program took about half an hour and worked first time - which effectively cured my interest in prolog, and proved my algorithm was right as far as I was concerned, I just couldn't write it correctly in prolog.

I used SWI-Prolog during an internship for the local city government over 20 years ago, we were developing an application where you could load a city map with streets and bus stops as a graph and it would determine the best routes within two points by car, bus or foot. Pretty much a primitive version of what modern map/GPS apps do. I don't remember which algorithms we used but it involved lots of backtracking, so Prolog was deemed the ideal choice for the routing engine.

It wasn't a pleasant experience, while I understood the fundamentals of the language quite well, once the dreaded "cuts" entered the picture it became impossible for me to follow the different paths within the code. In the end it kind of "worked" for bus and walking routes but the driving ones would almost certainly fail due to the one way streets. In the end it seems they got more talented/experienced developers to work on it since about a year after I left they released a somewhat decent version to the public in their website.

I haven't gone back to Prolog or declarative programming in general since then, it may be the "elegant" way to solve certain problems but good old imperative languages are much easier to maintain and debug.

Yeah well elegance doesn't matter to me. Removing this intrusive insect does.

It's somewhat comforting (and much needed at that) that newer versions of TP make more sense. I hope you were referring to 2.0.