Principles of Programming Languages (1997) [pdf]


99 points | by matt_d 34 days ago


  • siraben 33 days ago

    Denotational semantics are a great way to capture the meaning of a programming language. If your language provides a denotational semantics it's a nice exercise to implement them in a functional language. For instance, here's my implementation of R5RS Scheme in Haskell following the spec's semantics.[0]

    A lot of toy Scheme implementations omit call/cc for instance because it can be hard to reason about and do correctly, whereas following the spec made it really easy.


    • jpz 33 days ago

      "Denotational semantics are a great way to capture the meaning of a programming language" - big claim - what is the meaning of a programming language?

      For example, if you able to capture what the meaning of a programming language is, then could you please demonstrate by communicating what "the meaning of the programming language" C is, for instance? This appears to be a nonsensical combination of words, as I do not recognise a programming language as having an intrinsic meaning?

      • nv-vn 33 days ago

        I think you're thinking of the wrong definition for "meaning." OP is talking about what code in a language means, i.e. when I see the code say `1 + 2` does that mean that the numbers signed or unsigned? Are they arbitrary precision or a certain size? What errors can result from the addition? Etc. It's essentially saying that you can look at a piece of code and formally define what it means to the computer

    • p0llard 34 days ago

      This seems pretty good from a brief flick through; the author. Graham Hutton, is pretty well known in the CS community.

      Some other resources for people who find that this piques their interest and that they want to go deeper:

      * Formal Reasoning About Programs ("FRAP") by Adam Chlipala - available for free here: Forms the basis for 6.822 at MIT, and comes with an accompanying Coq formalisation (as well as psets in Coq).

      * The Formal Semantics of Programming Languages: An Introduction by Glynn Winskel.

      • siraben 33 days ago

        I'd also add

        * Essentials of Programming Languages by Friedman and Wand

        It's quite accessible and has an emphasis on executable implementations. Covers quite a wide variety of paradigms, FP, FP with mutation, type checking, type inference, and objects and classes.

        • tom_mellior 33 days ago

          Another good one is:

          * Concrete Semantics by Nipkow and Klein, which is also an introduction to the Isabelle/HOL proof assistant and formalizes all of the PL content in Isabelle

          There's also:

          * Semantics with Applications: An Appetizer by Nielson and Nielson, apparently a newer edition of an older book I liked and which used to be available as a free PDF, but the link from my browser history is now broken

        • matt_d 33 days ago

          General Program Analysis Resources may also be of interest:

        • dang 34 days ago

          If curious see also

          2017 (1 comment)

          The pdf is marked 2016 but probably dates from much earlier, so maybe it's best not to put a year in the title.

          • stevesimmons 34 days ago

            Late 1996 or early 1997, based on the PDF's reference to Hugs 1.3.

            The Hugs 1.3 manual is dated August 1996. Hugs 1.4 is from 1997. And then came Hugs98.

            • dang 34 days ago

              Nicely done. Added!

          • mathgladiator 33 days ago

            This is timely, I'm reviewing this deck just after working a grueling day on my language and tuning the language runtime for performance.

            What's interesting is how many things we can now intuitively take for granted, but for each idea someone had to suffer and make it real. Their primary barrier was people via communication and getting traction.

            • hn3333 33 days ago

              I like your second paragraph. This is true for large parts of life and a good reason to work hard and leave something behind for future generations.