I feel the exact same way. I’m grateful for modern computers and what they can do, but I think the substrates of Lisp and Smalltalk machines make building flexible component-based software easier than the Linux, Windows, and Web ecosystems we have today. If I had the spare time, I’d work on a modern-day OS inspired by the Lisp and Smalltalk environments of old.
If I had the time and the money, I’d like to pick up where Xerox PARC left off when they stopped working on Smalltalk, Cedar, Mesa, and similar projects. I’m also very fascinated by Apple projects of the 1990s such as SK8, Dylan, and the original proposal for a Lisp-based Newton. During the “interregnum” years at Apple many people with interesting ideas on system design and usability worked at Apple, such as Don Norman and Larry Tesler. I’m grateful for Steve Jobs’ return and for NeXT-based macOS, but unfortunately as time passed by, the Smalltalk, Lisp, and even NeXT influences at Apple faded away. It would be cool if somebody continued this vision. I’d do it in a heartbeat if I had the time and the financial resources.
You might want to check out Urbit: urbit.org. The whitepaper[1] is a bit outdated but in section 12 "Inadequate summary of related work" you can see some of its influences:
"Many historical OSes and interpreters have approached the SSI[2] ideal,
but fail on persistence, determinism, or both. In the OS department,
the classic single-level store is the IBM AS/400 [18]. NewtonOS [19]
was a shipping product with language-level persistence. Many image oriented interpreters (e.g., Lisps [20] and Smalltalks) are also SSI-ish,
but usually not transactional or deterministic. And of course, many
databases are transactional and deterministic, but their lifecycle function is not a general-purpose interpreter."
Urbit is mostly marketing fluff for extracting money from investors and potential users. It is based on a for-profit commodity and ecosystem (their stars and ships and planets and all that stuff) that you have to buy into to use it. You cannot just "run your own urbit".
Sadly, Urbit is forever tarred by one of its contributors. I'm not saying it's not worth investigating. It is. But why put your energies in a software project which already has a cultural strike against it? It's sort of like maintaining RieserFS: technically interesting but upsetting the social nature of programming humans.
Well. Rieser tried to keep his professional and criminal lives separate. RieserFS wasn't an expression of his revolutionary worldview.
And I was going to point out that Java survived Patrick McNaughton, FOSS survived ESR, etc.
From the whitepaper, Urbit looks bitchin. I love its audacity. In that way, reminds me of Linda (tuplespaces), Xanadu, Jef Raskin's Humane Interfaces (Canon Cat), and others.
Alas. It appears Urbit and its creator's worldview are inseparable.
I'll wait for the reboot. Or maybe just glean some of its ideas.
If I had the time and the money, I’d like to pick up where Xerox PARC left off when they stopped working on Smalltalk, Cedar, Mesa, and similar projects. I’m also very fascinated by Apple projects of the 1990s such as SK8, Dylan, and the original proposal for a Lisp-based Newton. During the “interregnum” years at Apple many people with interesting ideas on system design and usability worked at Apple, such as Don Norman and Larry Tesler. I’m grateful for Steve Jobs’ return and for NeXT-based macOS, but unfortunately as time passed by, the Smalltalk, Lisp, and even NeXT influences at Apple faded away. It would be cool if somebody continued this vision. I’d do it in a heartbeat if I had the time and the financial resources.