Just to argue the devils position … I think Alan Kay completely failed in delivering something novice friendly and commercial. All his research and work is moot and what is left is merely a pipe dream of personal computing. What worked in the real world was BASIC with its brazen gotos that helped some of the best programmers we know now, fall in love with programming. Forget about bicycles, BASIC gave creative minds jetpacks. It was ultimately Self and Javascript that succeed and not smalltalk.

Commercial... You know what's easy to sell? Drugs. What's hard to sell? Books. Especially physical ones, and technical books are particularly challenging. A dying business really. That's probably tells you something about commerce. 🙂

Economic constraints are important. I think books, papers are a terrible medium for technical communication. It’s one way and often preachy. In many ways the interactive pedagogy first seen in BASIC magazines was the future of programming education as opposed to the heavy ide driven approach or large technical manual approaches for better or worse.

There are pedantic, obvious, and correct reasons to disagree, but I'm here for xyzzy's perspective

What I take from "founding fathers" of our field these days is mostly their enthusiasm

xyzzy Interesting thoughts! I will try to privode soem related thoughts.

There is a interview with Adele Goldberg in which she points out how she and a few others went off from parc to build a company and to create something what today is known as open source

I found it quite interesting: Interview with Adele goldberg

https://www.youtube.com/watch?v=IGNiH85PLVg

here is my zettelkasten note for the interview:

• worked with Alan Kay at PARC
• interview on PARC industry, american relationship between academic research, industrial sponsered research and ,
• starting business into corporate buisiness sponsored by ven. cap.
• developement of open source as a social developement

Another thing you mentioned I see similarly critical, which is how alan kays work failed to have an impace in the broader culture of industry, education etc.. Of course its a complex Issue But lets take the STEPS project as an example: Here is the HN discussion thread about that report: Interview with Adele goldberg

https://news.ycombinator.com/item?id=11686325

I commented there with the user "notusinglinux" , I basically argue that to little effort was made to make the source of the project available, even if they didn't believe people would understand it. And IF they thought people would not understand it, why not create some kind of educational effort where to explore these ideas with a larger community? A MOOC or something related?

I tracked this quite a bit for example, a substantial part of the STEPS project was the graphics subsystem: NILE and GEZIRA. Dame Lang was the key collaborator here, he actually did publish some stuff on github and you can see how people were REALLY interested in that stuff. This is the relevant discussion ( you will find me there too):

https://github.com/damelang/nile/issues/3

people really tried to convince him that he should be working on this and explain it more to people, eventually Dan said : he would need about 6000$ a Month as compensation. And somehow this was the end of the discussion there. He never made a patreon account or opencollective or something other ( as far as I know).

Which leads us to the Funding problem and why I'm still interested in the Blockchain/web3 space, for example recently I came through the PKM topic to two open source Projects: athens research and logseq. Both are also struggling with funding, but at least have patreon and opencollective accounts. To move beyond this kind "we donate some money to our favorite OS project" another approach, even a different culture is needed. Experiments like these give me hope : https://radicle.xyz/blog/radicle-orgs.html

In lex fridmans recent interview with Jaron Lanier, Lanier mentions a combination of Tik tok and github. Spontaneous yet deep and valuable collaboration. https://www.youtube.com/watch?v=Fx0G6DHMfXM

Andreas S Yes funding issues could explain why Smalltalk could not succeed in education. It is certainly a complex project. Interestingly it seems the current generation of programmers learnt code via scratch, which is sorta inspired by smalltalk. Personally I grew up with Logo and Turtle Graphics. However since I read about the history of BASIC, commodore64, Amiga, Early Microsoft, Visual Basic 6 … I can’t help but think that BASIC was hugely important in influencing modern programming practices, well beyond what Smalltalk or daresay even Lisp ever did. In many ways BASIC is a dynamic version of Fortran. If you haven’t read https://10print.org/ you should check it out … its definitely eye opening.

I would even posit that Lua is the proper successor to BASIC … in its beginner friendliness, usage in games and Fortran heritage.

Smalltalk had and has some unique aspects, the deep embedding with the hardware the blurry line between OS and programming system.

HyperCard was also a interesting approach as people moved away from the potential concepts suggested from xerox parc the technology (computer software and hardware became driven by a kind of monopolistic industry culture). HyperCard was a apple idea/product but look how much creativity/agency/sovereignty apple tools give you today

One of the interesting concepts out of that STEPs/ VPRI context was that of virtual time. It actually made it into a JS lib: https://twitter.com/codefrau/status/1430715818115883008?s=20

Hypercard is wonderful. That reminds me, I have do my project of transpiling to lisp from a logo / newtonscript like syntax. I believe Hypercard also influenced javascript. I think the light weight approach was key for the success of javascript.

Thanks Andreas S for your summary of a part of history that I was mostly unaware of until now!

As for Basic, that's what I started computing with myself. Its single selling point was that it ran on machines people could actually afford. For me as a high-school student in 1982, it was Basic or nothing. Easy choice 🙂

With hindsight, Basic didn't really matter. For enthusiasts, anything is OK to get you started. It's a different matter for people who need a gentle push to get interested in coding. If many of today's programmers got started with Basic, that's probably survival bias to a large degree.

Super cute.

Love this

Nitpick but sqrt 36 is also -6. IMHO much tricker to resolve multiple paths with a linear structure..

It is a little too cute: needs to handle some ambiguity.

I want this but.. like.. with 1000 times more exploration and iteration. This is a good (if common) first step, at least.

Adding hierarchical type classification can result in programming communities that spend most their time talking about the "right" way to organise the universe of type classes, and experience surprisingly dogmatic discussions about that

I feel seen

FWIW I am also an opponent of hierarchy. It's one of those organizational concepts that has always sounded good in theory, but has always caused problems in practice. Is there even a single historical example of a hierarchical programming language concept making things better without any "gotchas"?

Of course the post is making more than just a point about hierarchy. It's opposing type-level programming.

cough Haskell cough OCaml cough

Adding type-level programming of any kind can lead to communities where the most empowered programmers are those with deep expertise in certain kinds of highly abstract mathematics (e.g. category theory).