3 ms·

> Lisp culture genuinely prizes getting-shit-done attitude. SICP tells me otherwise.

by anthk 2mo ago

vincent-manis 2mo ago
SICP is not intended as a how-to-program manual, it attempts to lay the foundation for a big chunk of computer science. “How To Design Programs” might be a better choice for someone learning Scheme. (But SICP will turn some of the people on to CS as a discipline.)
Jach 2mo ago
SICP, as beautiful and good as it is, is probably responsible for a large amount of the misconceptions and negative thoughts about Lisp. SICP isn't a Lisp book, it's a Scheme book, and the Scheme culture is at least as different to Lisp (Common Lisp and its ancestors all the way back to McCarthy) as Java is to Python.
- anthk 2mo ago
  I know both CL and Scheme, thanks. And both philosophies overlap a lot. Heck, there's an Scheme interpreter in PAIP as en exercise... I'd woudn't consider SICP negative, the same with all the good CL books (Intro to Symbolic Computation, Paradigms of AI Programming...) Both are the ur-examples on how to grasp the basic of CS well.
iLemming 2mo ago
Are you using a 42 year old book to find flaws in my argumentation? Let's talk real, contemporary numbers. NuBank: from ~12M customers in 2019, grown to 25M in a year, then to 48M at IPO a year later, then to 114M in 2024, to 131M in 2025 - 991.67% growth within 6 years and still going. Not in theory, not conceptually, not because "SICP has told them". They did get-the-shit-done. Not acknowledging that Datomic and Clojure has something to do with this succes would be simply a dishonest gesture. Walmart - the canonical "Clojure at scale" story. The eReceipts system processed every purchase across 5,000+ US stores plus online/mobile, built and maintained by just 8 developers. Architect Anthony Marcar's line after Black Friday: the system "handled its first Walmart Black Friday and came out without a scratch". They cited 5-10x less code than alternatives. Apple - known for using Lisp for a long time. Metabase - open-source BI. ~46k GitHub stars. Amperity - customer data platform, "99% Clojure" - their own quote. Netflix - been using Clojure since forever. Watch their talk from the last year's Conj. It's eye-opening on stability of complex systems. Cisco - malware analysis platform. Multiple huge Clojure projects: IROH - Incident Response Orchestration Hub, CTIM/CTIA - Cisco Threat Intelligence Model and API, etc. Braintree/PayPal - payments backend. CircleCI. Grammarly (this guys are on SBCL) And that's just a short list of actual, real, profitable businesses built with and maintained using Lisp. I don't know what "SICP has told" you, but maybe you just need to look around, things have changed a bit since then.
- anthk 2mo ago
  I mean, both Scheme and CL have good libraries and a set of standards.
  - iLemming 2mo ago
    Okay, but what that has to do with preferring strong CS theoretical grounds vs. cutting corners to get-the-shit-done? My initial point was about cultures - Haskelites are typically very smart, extremely mathy, and they'd often choose certain ways even if it takes them forever and requires reading and analyzing dozens of academic papers, just because "Galois was a genius and we can make this theory fit here, because then we can center infinite divs on infinitely large DOM entities...", then publishes a paper "Coequalizers for Vertical Alignment: A Comonadic Approach" A Lisper would be like: "Ahem... I wrapped the whole thing in a macro. Yes, it evals a string. Yes, the string is generated by another macro. No, I will not be explaining it. It shipped Tuesday, it prints money, it works. Gotta go, kids need dinner.."
    - anthk 2mo ago
      You forgot about HAKMEM under PDP10 assembly/ITS/MacLisp/Macsyma (these three/four ovelapped)
    - rootnod3 2mo ago
      So, Haskell people smart, other people dumb? Also, where do you go from "preferring strong CS theoretical grounds" to "cutting corners"? They don't have to be mutually exclusive. That garbage collection that Haskell has? Guess where that came from. The whole "functions as data to pass around", guess where that came from. It was those pesky lispers on a Tuesday. Edit: typo
      - iLemming 2mo ago
        > Haskell people smart, other people dumb? Da hell you're talking about, friend? I said nothing about "people", or them being smart or dumb; just a difference in cultural stereotypes. Now you're ironically "proving the point" I didn't make by saying some dumb shit and pulling me into this pit too. If I was actually smarter I'd probably just ignored it. Reminds me Billy Madison quote: "everyone in this room is now dumber for having listened to it. I award you no points, and may God have mercy on your soul."

> Lisp culture genuinely prizes getting-shit-done attitude. SICP tells me otherwise.

by anthk 2mo ago

> Lisp culture genuinely prizes getting-shit-done attitude. SICP tells me otherwise.

SICP is not intended as a how-to-program manual, it attempts to lay the foundation for a big chunk of computer science. “How To Design Programs” might be a better choice for someone learning Scheme. (But SICP will turn some of the people on to CS as a discipline.)

Jach 2mo ago

SICP, as beautiful and good as it is, is probably responsible for a large amount of the misconceptions and negative thoughts about Lisp. SICP isn't a Lisp book, it's a Scheme book, and the Scheme culture is at least as different to Lisp (Common Lisp and its ancestors all the way back to McCarthy) as Java is to Python.

anthk 2mo ago
I know both CL and Scheme, thanks. And both philosophies overlap a lot. Heck, there's an Scheme interpreter in PAIP as en exercise... I'd woudn't consider SICP negative, the same with all the good CL books (Intro to Symbolic Computation, Paradigms of AI Programming...) Both are the ur-examples on how to grasp the basic of CS well.

iLemming 2mo ago

Are you using a 42 year old book to find flaws in my argumentation? Let's talk real, contemporary numbers. NuBank: from ~12M customers in 2019, grown to 25M in a year, then to 48M at IPO a year later, then to 114M in 2024, to 131M in 2025 - 991.67% growth within 6 years and still going. Not in theory, not conceptually, not because "SICP has told them". They did get-the-shit-done. Not acknowledging that Datomic and Clojure has something to do with this succes would be simply a dishonest gesture. Walmart - the canonical "Clojure at scale" story. The eReceipts system processed every purchase across 5,000+ US stores plus online/mobile, built and maintained by just 8 developers. Architect Anthony Marcar's line after Black Friday: the system "handled its first Walmart Black Friday and came out without a scratch". They cited 5-10x less code than alternatives. Apple - known for using Lisp for a long time. Metabase - open-source BI. ~46k GitHub stars. Amperity - customer data platform, "99% Clojure" - their own quote. Netflix - been using Clojure since forever. Watch their talk from the last year's Conj. It's eye-opening on stability of complex systems. Cisco - malware analysis platform. Multiple huge Clojure projects: IROH - Incident Response Orchestration Hub, CTIM/CTIA - Cisco Threat Intelligence Model and API, etc. Braintree/PayPal - payments backend. CircleCI. Grammarly (this guys are on SBCL) And that's just a short list of actual, real, profitable businesses built with and maintained using Lisp. I don't know what "SICP has told" you, but maybe you just need to look around, things have changed a bit since then.

anthk 2mo ago
I mean, both Scheme and CL have good libraries and a set of standards.
- iLemming 2mo ago
  Okay, but what that has to do with preferring strong CS theoretical grounds vs. cutting corners to get-the-shit-done? My initial point was about cultures - Haskelites are typically very smart, extremely mathy, and they'd often choose certain ways even if it takes them forever and requires reading and analyzing dozens of academic papers, just because "Galois was a genius and we can make this theory fit here, because then we can center infinite divs on infinitely large DOM entities...", then publishes a paper "Coequalizers for Vertical Alignment: A Comonadic Approach" A Lisper would be like: "Ahem... I wrapped the whole thing in a macro. Yes, it evals a string. Yes, the string is generated by another macro. No, I will not be explaining it. It shipped Tuesday, it prints money, it works. Gotta go, kids need dinner.."
  - anthk 2mo ago
    You forgot about HAKMEM under PDP10 assembly/ITS/MacLisp/Macsyma (these three/four ovelapped)
  - rootnod3 2mo ago
    So, Haskell people smart, other people dumb? Also, where do you go from "preferring strong CS theoretical grounds" to "cutting corners"? They don't have to be mutually exclusive. That garbage collection that Haskell has? Guess where that came from. The whole "functions as data to pass around", guess where that came from. It was those pesky lispers on a Tuesday. Edit: typo
    - iLemming 2mo ago
      > Haskell people smart, other people dumb? Da hell you're talking about, friend? I said nothing about "people", or them being smart or dumb; just a difference in cultural stereotypes. Now you're ironically "proving the point" I didn't make by saying some dumb shit and pulling me into this pit too. If I was actually smarter I'd probably just ignored it. Reminds me Billy Madison quote: "everyone in this room is now dumber for having listened to it. I award you no points, and may God have mercy on your soul."