<https://dreamsongs.com/WIB.html>
That's a thirty...
# linking-togethera
Alex Chichigin
08/20/2021, 10:07 AMhttps://dreamsongs.com/WIB.html
That's a thirty year old (😱) piece and most likely familiar to many members of our community, but it still has some very interesting and relevant insights on design philosophy, programming languages and IDEs. In particular, it introduces a notion of program language as opposed to programming language, a sort of metalanguage.
k
Kartik Agaram
08/20/2021, 3:32 PM
Which parts? I like RPG's writings, but with this one I've always had a hard time ignoring all the AI-winter stuff.
Kartik Agaram
08/20/2021, 3:34 PM
Ooh, this is an interesting statement I hadn't noticed before:
continuations remain an ugly stain on the otherwise clean manuscript of Scheme.
Kartik Agaram
08/20/2021, 3:39 PM
It's not clear to me why he cares about continuing the Common Lisp standardization process if most of section 3 is about the successor to Common Lisp. Though this too is about the time the article was written in, so not that important.
a
Alex Chichigin
08/20/2021, 4:31 PMMost relevant probably the parts about IDEs.
The part about "the right thing" vs. "the worse is better" and how to get best of both worlds is pretty interesting to me too.
Comparing his points about "future Lisp" with Clojure "the Lisp that could" is pretty curious. Also I'm thinking about Racket which kinda successfully lives in a sorta parallel universe.
Alex Chichigin
08/20/2021, 4:33 PMAs long as the article was written even before ANSI Common Lisp, finishing standardising that was paramount. Then making it international, which didn't happen as a standard AFAIK. After that it starts to make sense to devise a "better Common Lisp". Which again never happened, and I'd be curious to know why exactly.
Alex Chichigin
08/20/2021, 4:34 PMBesides I was surprised how much AI winter influenced Lisp.
Alex Chichigin
08/20/2021, 4:41 PMAlso,
The largest criticisms of Symbolics in the article are that Symbolics believed AI would take off and that Symbolics mistakenly pushed its view that proprietary hardware was the way to go for AI.It's kinda funny how relevant it sounds. 🙂
k
Kartik Agaram
08/20/2021, 5:35 PM
Oh, I love "worse is better". I forgot that this is the article that has it. FYI my favorite summary of "worse is better" (also by RPG) is in page 219 of https://www.dreamsongs.com/Files/PatternsOfSoftware.pdf:
It is far better to have an under-featured product that is rock solid, fast, and small than one that covers what an expert would consider the complete requirements.
Kartik Agaram
08/20/2021, 5:55 PM
By "parts about IDEs" do you mean 3.2?
Kartik Agaram
08/20/2021, 5:58 PM
As long as the article was written even before ANSI Common Lisp, finishing standardising that was paramount. Then making it international, which didn't happen as a standard AFAIK. After that it starts to make sense to devise a "better Common Lisp". Which again never happened, and I'd be curious to know why exactly.The process of standardizing Common Lisp was so arduous was that the people who migrated in to it have forever been hostile to anybody and anything outside it. It's like, "I just finally got all moved in. What do you mean you want me to move out? And you're trying to poach my neighbors?! Enemes FOREVA!" Maybe I have the benefit of hindsight, but it doesn't make sense to gate something on a standard if it's going to be incompatible with it anyway.