Fun shouldn't be a goal — it should be a consequence of something stimulating (whether driven extrinsically, like instruction, or intrinsically, like self-discovery — either of which can be effected by play) in the absence of https://blog.nindalf.com/posts/factorio-and-software-engineering/https://blog.nindalf.com/posts/factorio-and-software-engineering/.

Wow I'm surprised to see you link to Factorio! The kind of flow state that game induces is exactly what I want programming to feel like (post requirements analysis). Factorio provides a visceral feeling of progress. You're able to see how every addition you make affects your system, you're able to assess its internals, and you're able to make a plan for your next step, and execute it straightforwardly.

I don't think the "conveyer belts" metaphor translates though. Programming is never really conveyer belts (transfer of goods), it's more like a Rube Goldberg device (chain reactions).

I think I agree with this, personally. I am much happier working forwards from something I completely understand, than trying to make sense of something that already exists.

This is the same motivation that leads people to use Excel for everything, often from scratch every time.

Sometimes I think that this is the primary reason why software reuse hasn't succeeded as much as some people would have wanted to. Not because OOP is flawed or whatever, but because people... just wanna have fun 😛

Easier, but not necessarily optimal. The distinction seems related to the one between ease and simplicity. There's a sense in which learning a CAD program is simpler than building one yourself, but whether it's easier depends on your starting point. Path dependence just ruins everything. :D

Learning to model in 3D is hard, as in expect to spend 2-3hrs a day for 15 to 30 days to learn Blender or Maya or 3DSMax or Cinema4D or Houdini or ZBrush or AutoCAD and maybe someday someone will figure out a way to make it easier (VR?) but when you look into it there's a reason and purpose for every one of the 6000+ features of blender or maya or 3dsmax and 2 weeks of work might get you some cubes and spheres on the screen but you won't be modeling characters for Uncharted or sofas for Ikea or cars for BMW.

It's nice if they are having fun but the idea that they got frustrated spending 30 minutes in a modeling tool and then spent 2 weeks making their own suggests mostly that in a year or two they'll realize they really should have spent the time to learn a modeling package if their goal was to model something.

It mean it's kind of like saying "I spent 30 minutes trying to learn guitar but it was too hard so I wrote some code to make chiptunes". Yea okay but you still don't know how to play the guitar and no amount of diddling with chiptunes is going to get you closer to being able to play that guitar.

PS: there are programs that make 3D modeling easier but they almost always fall short once you get serious. Like notepad is easier than vi and iMovie is easier than Final Cut Pro and MSPaint is easier than Photoshop

➕ the talk references Bret Victor's "Future of Programming" 😉

Inspiring talk! Thanks for sharing!

Regarding pre-VoodooPad personal Wikis, a quick nod toward 2002's Tinderbox: https://en.wikipedia.org/wiki/Tinderbox_(application_software)

... which was part of a wave of notetaking apps and personal knowledge bases in the early 00s. Those interested in this sort of thing should take note of the graph view, the bidirectional links, the expandability, and so on, as covered here:

http://www.atpm.com/10.10/atpo.shtml

Absolutely, I remember that app well, I really miss those days when so many new ideas were being explored

For a while I thought this was the case too, but I don't buy into this narrative any more. There is a ton of innovation happening in smaller "scenes" that just isn't very legible to media outlets and these influencer types and VCs. Imo the most interesting work is often the most obscure. The frontier is not glittzy and has no press-releases, in fact often it might be deliberately staying under the radar. So I think the opinions about innovation from these established personalities should be taken in mind with the maturity of their careers, which often corresponds with how deeply embedded they are in stale systems.

I don't disagree. (I'd love to hear any examples of innovative work.) Based on your description, it's unclear if the innovative work you're talking about has also been market viable? That was a nice thing about the early 2000s, many of these products were also profitable.

yoshiki So what/where are those small scenes, and how would an outsider find them? I think discoverability matters as well for innovation, because it's a condition for getting feedback from diverse sources, and not just a small bubble of friends.

Konrad Hinsen I'm still getting familiar with it but the work that the p2p community is doing is incredibly imaginative. https://n-o-d-e.net/zine/ has a good overview of their efforts.

yoshiki Thanks for that pointer! I hadn't seen it yet, although I do follow the p2p community a bit, my contact point being IPFS. There is indeed a lot of creative work going on there, coupled with serious efforts to make p2p usable even within today's world dominated by centralized systems.

I agree that discoverability matters. I think it's a very subtle issue tho. I wanna write up some of my thoughts on it soon, since I feel like I've learned a lot about this in the last year or so.

Certainly communities like this help tho!

holy crap! thats amazing

apparently Playmaker was also used as part of makinng INSIDE

which is one of my fav games

Ooh, check out https://comby.dev/docs/overview

Also I can recommend Hammock driven development. It is not strictly about creating programs, but more about how a person can create something big and elegant. This talk really blew my mind, I watched it 2 times in a row making notes 🙂

Your list is good btw, really like those talks

I've always disliked the for loop, and it took me forever to realize that the term comes from set comprehension notation in Math. Great to see it replaced here.

I’ve been a keen follower of Casey and his handmade hero series for a while, along with his partner in crime, Jonathan Blow. Their rants are always fun, and I agree with most of it. His recent demo of how Visual Studio is slower than the version from 10 years ago was telling. The statement that software which works in the morning isn’t guaranteed to work in the afternoon is frightening and so true! A feature I’m keen to build into my own software is a snapshot ability to enable retreat to any modification point. Perhaps that’s something that should happen at an OS level? Because I can’t see a day where we don’t need updates to patch vulnerabilities, if nothing else....

Yes, of course snapshots should happen at the OS level. This is decades overdue, but we are seeing progress in the Linux world. Ubuntu 20.04 has ZFS on root, with automatic snapshots when a package is upgraded. Fedora 33 (due Oct 2020) defaults to BTRFS on root, again with automatic snapshots and the ability to roll back to a previous state. NixOS has a different approach to package management, allowing individual apps to be rolled back, rather than requiring rollback to an earlier state to be a "global" operation. The flatpak app distribution system (works on all modern Linux systems) permits rollback to a previous version (which is granular on a per-app basis). What seems to be largely missing so far is a slick, high level GUI for rolling back software changes.

I have been using https://guix.gnu.org/ (which is based on the same ideas as Nix) for a while to stabilized my software environment, and I wouldn't want to live without it. You can not only snapshot your installed software, but even install the state from an arbitrary point back in time (though that can be slow because everything gets recompiled, starting with the compiler).

GUI support is indeed missing, and some important software hasn't been packaged yet. You can't install binaries made for ordinary Linux distributions either, because the file system layout is very different. So this is not ready for prime time, but it's far more than just a proof of concept.

I really wish this future would happen, but we’d need someone with a ton of money to do it and I think that’s very unlikely anytime soon. It always felt like hardware interfaces should be implemented as a blackboard system where devices push events and consume events and nothing else (similar to his “just have ring buffers” idea). Interop is trivial at that point. Creating new devices is just pushing new events that some other system looks for. The events can be easily versioned and so on. The problem is that doing so isn’t just a ground up software rewrite, it really is a ground up hardware ecosystem reconstruction and boy is that a tall order.

(Aside: glad to see Nix and Guix getting some love in this thread)

What I find interesting about this beyond the immediate impact on the Python data science community is that it's the first standardization effort I am aware of that is run exclusively by Open Source communities. Some of them are backed by for-profit companies, and that may make a difference, but it's still an interesting first.

I wonder if @Paul Ford still visits the Slack...

Yep, totally agree that manipulating data is where our contribution would likely be most impactful for end users. I’ve been on a https://stefan-lesser.com/2019/12/06/structure-and-behavior/.

I read this as Notion and was like, “It’s fine, I guess, but breakthrough seems like a stretch”

Well I'm missing context to judge the impact.

I wonder if there is something similar with code. Also looks like what it was, was a different UI to underlying equations.

Physicist here. The claim is not exaggerated: Feynman diagrams have changed the way quantum physics is done. Think of them as a layer of abstraction on top of mathematics. Each diagram can be translated to some lengthy integral, but those integrals are much too lengthy to be manipulated in all but the simplest cases.

BTW, another, more accessible, case of notation being a major breakthrough is vector notation in geometry. Check out old textbooks (1930s or earlier) in mathematics or physics - you don't want to look at those equations if you grew up with vector notation.

Btw sounds like a #thinking-together post :)

Ope Hmm… was thinking that too, but then decided that I'm basically sharing and asking for links. Well, now it's here; let's see how it goes.

I guess https://github.com/akkartik/mu from Kartik Agaram is built around the same idea.

I like this one from a low level hardware point of view: https://www.youtube.com/watch?v=HyznrdDSSGM (I have built some of this one myself)

And this one: https://www.nand2tetris.org/

But neither of these go as far as making an OS.

Robert Butler's https://github.com/grokthis/ucisc

Smalltalk, as discussed by Dan Ingalls: https://www.cs.virginia.edu/~evans/cs655/readings/smalltalk.html. That's not how Smalltalk turned out to be used, but the idea was there.

Not sure if Knuth's Mix counts, given its pedagogical focus: https://www-cs-faculty.stanford.edu/~knuth/mmix.html

https://collapseos.org/

Seems like https://urbit.org/understanding-urbit/urbit-os/ fits that description too.

And then I learned that Project Oberon has a successor: https://en.m.wikipedia.org/wiki/Bluebottle_OS

Xv6, a simple Unix-like teaching operating system:

https://pdos.csail.mit.edu/6.828/2020/xv6.html

https://en.wikipedia.org/wiki/Plan_9_from_Bell_Labs is actually quite self-contained and easy to understand.

DreamOS (scheme all the way down)

http://web.archive.org/web/20120815232133/http://www.stripedgazelle.org/joey/dreamos.html

ArmPit Scheme for ARM-based MCU

http://armpit.sourceforge.net

MiniOS

https://arxiv.org/abs/1811.09792

i really like the phrase “public luxury”—communicates much better the value of things like libraries, parks, the postal service (IMO)

This highly personalized model of social change proposed an individual solution to a structural problem, which necessarily neglected the wider social context.Yeah I've been thinking about this exact thing lately as well. Open-source has historically lacked intersectionality with other, more obviously material social issues. FOSS's roots in the problematic and esoteric "hacker ethos" also doesn't help.

I am optimistic about the future of open source though. I think many people are recognizing right now, including of course the author, that GNU-style open source alone is insufficient, that it needs to be placed in and complement larger liberatory efforts. Have you checked out the solarpunk movement? Open source features heavily, but as I understand it is more of a derivative effect of a more fundamental belief in mutuality and autonomy.

I agree with yoshiki. There is a hard core in the FSF that still promotes GNU-style freedom as the ultimate road to happiness, but most people in the Open Source universe have recognized that licenses aren't the whole story.

My main worry about the FOSS world right now is a tendency towards tribalism in Open Source communities. For many people, their community is everything, more important even than the software it produces and the impact it has on the wider world. As an example, for many people in the scientific Python ecosystem it seems more important to promote Python than to enable good science.

yoshiki solarpunk is interesting, as a cultural and aesthetic movement I think it serves to reclaim the future, at least one possible future. What I mean, is that we have few large sources of inspiration for our future as an aesthetic: the techno-determinist futures found in shiny Apple products, Elon Musk and platforms like Uber; and the reactionary futures as a response to climate change.

I think both of these are problematic and not actually desirable (even as I write this on an iPhone, which I do think is quite sexy)

I love any effort to rekindle a sense of what’s possible, especially building new and desirable aesthetics around it. and for anyone familiar with political theory this is all closely related to ideas such as ‘capitalist realism’, the ‘end of history’, and ‘modernity’

Konrad Hinsen totally agreed, and your example of Python vs science is a good and worrying example.

Fostering a sense of community is probably necessary to get people to devote the kind of energy a project like Python requires, though.

I think this take on open source undervalues many ways in which that movement increases the personal freedom and autonomy of users. Being able to fix things oneself (in the spirit of Right to Repair laws), audit one's system to make sure it's not running spyware, &c, are very good things that fall outside of the scope of this piece's focus on the means of production. (Which is not to suggest those questions aren't important!)

Jack Rusher I suspect the author would totally agree with you, she writes about those same feelings in a book called https://repeaterbooks.com/product/abolish-silicon-valley-how-to-liberate-technology-from-capitalism/ (about her disillusionment with the status quo, which might resonate with some here)

Chris Knott Definitely. Like for all large-scale coordination of human efforts. But then we get into the realm of https:/www.jerrypournelle.com/reports/jerryp/iron.html

, and with further growth the organization becomes a danger for its environment. It's the human version of the paperclip maximizer. It would be great if FOSS community leaders were aware of this mechanism and took preventive action, but so far I don't see that happen. I'd love to see codes of conduct also mentioning the community's impact on the outside world, and larger communities nominating outside observers to provide them with feedback.

Among the biggest damage FOSS communities can create is depletion of the global attention reservoir. Which I think is one of the issues of FoC work. Anyone doing serious work in tech nowadays has to delve so deep into Linux that it becomes hard to envisage computing without Linux.

No experience with Godot (yet), but...

Moving towards more integrated graphics stuff seems to be in the future of gui stuff

100% agree!

Totally worth trying out Godot. The tiny download size and (relatively) instant startup time makes it frictionless to quickly try. It’s kinda mind boggling if you’re used to using Unreal.

Interesting. That's pretty much how my own toolbox has evolved as well, though I wasn't really aware of it before reading this essay!

My tech stack is Emacs plus a few add-ons (org, org-roam, notmuch). I have an org-roam database for long-term notes, and two plain org files (work/personal) for short-time notes and to-dos. A crucial feature for me is that I can link from those two org files to my long-term notes, to e-mails, and to the individual files and directories in which I manage larger projects, in particular collaborative ones. Each project also has an org file for project-specific short-term notes, for the notes I share with my collaborators.

This is great, thanks for sharing. Now we only need to make it happen that everybody can do that, not just us geeks who know how to program.

Similar: https://futureofcoding.slack.com/archives/C5T9GPWFL/p1583608988119700

Reminds me of https://moldover.com who builds his own guitars.

Ei Wada is a hero. Makes me nostalgic for https://ivanish.ca/performance/ time playing in the noise / sound-art scene.

Good mini-doc about his work: https://m.youtube.com/watch?v=bQ97ujOAj6g

Delighted to see this show up in my player this morning.