Paul Tarvydas 2022-10-24 03:11:40
Nicholas Yang 2022-10-24 03:21:46
Paul Tarvydas 2022-10-24 03:44:03
Paul Tarvydas 2022-10-24 04:03:26
Future of Coding History
Does anybody know if there’s a parser generator like tree-sitter, i.e. produces a CST instead of an AST, but that can be used as a compiler frontend? Tree-sitter kinda can be, but it’s not the easiest
I have been using Ohm-JS for a year or two. I would be happy to discuss in more depth. I love PEGs and Ohm-JS in particular.
Had a daydream of an alternative reality where 1% of VC funding that goes into no-code had to fund this community instead, no strings attached. Didn't get to the part of how that'd actually work given the messy nature of people, but sure would be an interesting situation 😄
Where's all the ambition, the sense of adventure, that drove funding in the 60s? A time where the mother of all demos was possible to achieve with a small team that thought outside the box. Have VCs, researchers and funders become disillusioned. Have marketing teams cried wolf too many times. Have people given up on solving the really hard problems at the core of code. Who's picking up the mantle from the Engelbarts and Victors? Who's willing to fund their work and have it become products that changes people's lives.
I would love to live in that reality.. but at least in here in the Netherlands we have a special subsidy from our government which helps a little bit.. for me this means that I can work one full day a week on my project, which is better then nothing (subsidy comes in form of a tax reduction and since I do freelance work, that's really helpful)... I would offcourse love to do this fulltime😊
Ah cool, didn't know that NL has that. We're currently backed by the Danish Innovation fund, roughly 3.5K euro per month as salary over 12 months. Really difficult to be accepted into their program, but still, great to live in a country where that's even an option.
In my case, since I have a "one-person"-business ...I can deduct 12k from taxes on a yearly basis using that subsidy (wbso it is called). You need to apply for it, which is not that hard with a future-of-code project. It's also needed that you keep track of your spend hours (but not in much detail) which I am used to anyway, so not a big deal for me. 3.5K per month doesn't sound bad either
I'm so fortunate in this regard it's genuinely embarrassing. There are people inside the central agencies of the Government of Canada facilitating experiments into Rules as Code across the organization, and they have seconded me from my own consulting firm to work full time on open source Rules as Code tools for their projects. It won't last forever, I don't think, so I intend to push things as far as I can in the meantime.
I do think that we have to remember that there was geopolitics at play in how some things used to be funded. A lot of early AI research was a militaristic reaction to the 5th gen project in Japan, for example. If we are less afraid of one another as nations in 2022, that's probably a good thing. But we do need to find other ways to push the boundaries in public, not exclusively private, interests.
The Mother of All Demos emerged from a culture that was focused on value creation rather than value capture. VCs, "products", and "marketers" are all the opposite of that, which is why that work was funded with public money.
VCs are looking for ideas they can believe can work because something close enough also worked. Unfortunately, it's going to be difficult to get them to diverge from that, unless there's a hype craze like the recent crypto/web3 one. The closest success right now to the ideas here is probably Replit - though it isn't yet huge, it clearly has a strong product velocity and is something of a darling (and the sponsor of this podcast). Even new languages like Unison and Dark have been funded, but they both still look fairly similar to those in use now, and don't address end-user applications. Now, the narrative has shifted to AI, and I think any project here seeking funding will have to answer to that as well. VCs at least know that "worse is better" and successful innovation requires compromise - all else is storytelling.
Any insights as to why there's been a fundamental shift from a culture of value creation to value capture? Code has so much potential for good, but investment in better tools for coding and new paradigms seems super niche and limited.
Jim Meyer There was a general cultural shift in the Anglophone countries during the 80s (see: Reagan/Thatcher). A great many unfortunate downstream events have followed...
Jim Meyer Commercial tools and languages failed in the 90s, with the rise of the Web. They were too tied to desktop applications, and relied on expensive licensing strategies that guaranteed freely available languages would win, even if they were significantly inferior in quality. To Jack Rusher’s point, ARPA funding dried up in the 70s and completely closed off in the 80s, pushing the responsibility for development to industry, academia, and hobbyists, none of which could deliver on a cohesive vision of a programming environment like PARC managed to.
It's a challenging turn of events for sure. It's like the 60s and 70s thought of code and coding tools as critical infrastructure that everyone needs access to. I guess the problem is that takes it into political territory, and then it becomes a discussion around the degree to which it needs to be privatized. Everyone agrees (mostly) that roads, police, fire dept. etc. is paid for and used by everyone. Beyond that it's a divisive mess.
I think it was Alan Kay who had a talk where he said something along the lines of "technology and innovation needs people working on 10 year horizons". Today's technology world is obsessed with quarterly earnings and growth.
FWIW: A group of VC’s asked me to go and talk to Bret about Dynamicland. At that time they were running out of funds, and Bret was trying to figure out alternatives. My suggestion to him was threefold:
The problem is that there is a big difference between a demo and a product, and in many cases people want to build the former while others want to fund the latter.
The mistakes of more extreme political positions regarding innovation stem from an absence of ecosystem thinking. You don't get good discoveries without public investment in research, and you don't get refined products without a healthy private sector. Brad Myers has a very good paper on the research side from the 90s, some pullquotes and a link can be found here:
twitter.com/jackrusher/status/1552970752487559169?s=20&t=EXVojBXTF9dLLD7EOmDdQA
🐦 Jack Rusher in NYC: Brad Myers also wrote, around this time, an absolute banger on the history of HCI research. These picture-quoted bits from the intro of that paper are only more relevant today. /cc @Ben_Reinhardt @sarahcat21 https://www.cs.cmu.edu/~amulet/papers/uihistory.tr.html
Agree completely.. the trick IMHO is to build a clean bridge between the two..
99% of humans think of code as a medium that is too difficult. I don't think the medium is the problem. I think it's the way we've been asking end-users to interact with it:
So far, 99% of human-to-code interactions have been code-as-text, an interaction paradigm with unforgiving syntax and symbolic mental models.
Investments in interaction paradigms for code have almost exclusively focused on code-as-text and code-as-export, but there are other more end-user friendly ways of interacting with code that we can explore.
Let's break down the interaction paradigms for code (tell me if I missed any!):
There's so much unexplored territory outside the traditional code-as-text and code-as-export interaction paradigms! Would love to know what you think and whether I've missed any 😄
Under this framework, I guess I'm interested in code-as-executable. You do a thing, then you step back from the action, observing it, abstracting so as to apply in a new situation. Not exactly programming by example.
