Ask HN: What's Prolog like in 2024?

If I want to find something in the brain but not in bone structures. If I want to find something in a kind of cell but that have a nucleus.

They are also extremely useful for automated annotation. Your automated system may annotate with a upper term because it doesn't have enough information to be more precise. That's already a big help for a human to come and put a more precise term.

We are at a convergence of technologies, with ontologies, graphs, llms and logic programming. A lot of people were too early on this and discouraged from pursuing further by people that couldn't grasp why it was so important.

I don't think so. In your comment above you keep saying that it's not about syntax, then immediately after you switch to explaining how it really is. Notice this for instance. I say:

>>> Knowing Erlang will not help you learn Prolog just because the syntax looks similar.

You say:

>> That is not what is being suggested.

Then you go on to say:

>> But syntax is the gateway into the study of computer languages i.e. you first show the syntax and then explain its semantics. Humans learn new things using analogies/similes/metaphors/etc. Here similar syntax is a great help since it eases the cognitive load while learning new concepts.

And:

>> Start with your known syntax in a language you know, see how it maps to the same/similar syntax in the language you want to study and then lookup/compare/contrast the semantics of it in both the languages. It is like a cheat-sheet while studying a proper programming language textbook.

So it's hard for me to see how you're not suggesting that syntax will help you learn Prolog, when you're arguing that syntax helps you learn a new language.

And what I'm saying is that knowing something about the syntax of Erlang will not help you learn anything about Prolog. I don't think you can accept that at all, but it's true and it's my experience of many years watching people, including myself, trying to learn Prolog using their knowledge of other languages, either their syntax or semantics. My experience includes working as a teaching assistant for a Prolog course during my PhD. The truth is that Prolog is not an easy language to learn because it works very differently than any other programming language and you won't be able to use your knowledge of any other programming language to learn it. If you want to learn Prolog you have to start by finding a way to set aside most of what you know about programming. That is very hard to do, and so it's very hard to learn Prolog, except at a very superficial level.

I appreciate that's not easy to accept without long time experience.

But I don't think we're getting anywhere with this conversation so I'll respectfully bow out and thank you for the patient and civil exchange of views.

To settle this once and for all; in your previous post you gave the example of assignment which actually proves my point that you can easily go from Prolog to Erlang. Why? Because the behaviour is the same in Erlang i.e. there is no assignment but only pattern-matching. You can open a Erlang shell and type in X=a, X=b and it will barf. To show it even more convincingly type in Y=1+2, Y=3 and Y=4 and it will barf only on the last line. Lhs=Rhs are pattern-matched and variables are single assignment. A person with prior exposure to Prolog already knows this and hence can easily map it to Erlang. Only people coming from imperative languages (eg. C) need to adjust their mental model of the "=" operator. When i started with Erlang years ago that was what i had to learn first but thankfully all Erlang books cover it in the introductory chapter itself. Also functions in Erlang are written as a series of clauses which are pattern-matched.

I presume you do not have experience with Erlang and so your experience with only Prolog blinds you to the similarities. I highly recommend you get a Erlang book (Joe Armstrong's book is a good one since he uses much of the terminology like "terms", "clauses" etc. from Prolog) and with the help of the cheatsheet just try out the basic syntax in the Erlang shell and see what maps to Prolog. I am quite sure that you will find it convincing.

Finally w.r.t. Prolog's computation model being based on Predicate Calculus/First-Order Logic and so the model/semantics are very important is not being denied. I myself came to Predicate Calculus from the "Program Correctness" viewpoint of Floyd/Hoare/Dijkstra and hence an quite aware of its intricacies. But for most programmers without such a background in Set theory/Relations/Logic model building takes time and so syntax is their entry point and understanding happens over a period of time with study and trial-and-error.