Move to build

To me, the difference is that unlike the brownstones (in Brooklyn, Boston, and Philadelphia, each with their own twists - depending on the city), these are everywhere; there are so many more major cities in America, and architects can design not just for the city, not just for the nation, but often for the world. Homogeneity is beautiful when it's part of a central plan or culture for an area, but to globalize this kind of style is distressing!

This rich, written, familial history is beautiful - I hope I can motivate my parents to pursue this and encourage my descendants to do the same.

Wow - how was this orchestrated?

This dialogue has continued for as long as we've built buildings - to what degree do we give skilled craftsmen the challenge or autonomy to craft elaborate fixtures on their own vs. focusing on large forms declared by the architect that can be crafted with unskilled labor?

This relationship - this interface - between the architect and the construction work is fascinating to me. I love watching construction workers work. I love seeing progress. Why am I not one?

aha

Why not?

Is it?

I would love to know what such an evaluation looked like! How do you test for spacial awareness? This reminds me of that Blokus or Tetris game I played as a child that required fitting items into various spaces. I wonder often why I didn't bother studying architecture after that.

How do we 'de-scale' or 'de-industrialize'?

My guy. king

Food tastes so much better when it comes from someone who cares

Folk craftsmanship - caring about both the designer and the manufacturer. Reminiscent of Benjamin Woodward's work on the Oxford University Museum of Natural History - giving the craftspeople building the museum the opportunity to design the capitals as they desired.

They were contemporaries, but to what degree did they parallel one another? My current understanding is that VkHuteMas was centralized, communist, disciplined, to find and put forward propaganda, while the Bauhaus was at first an anti-establishment organization, standing against traditional architectural practices.

Agreed. Crawling 'forked' the web experience for many people; the crawlers see one thing, search engines preview another, and user clicks open a third, eschewing standards to 'optimise' for each of these three scenarios.

Ideally what is best for the user follows HTML standards and lends itself to the other use cases. Unfortunately, buzzwords have been overused to farm for clicks and ads, so this is not the case.

Why would this be desirable?

This is how I talk to my family, or used to - my dad's side in particular - and I had a bit of trouble learning to work with others and translate (I still do, to a degree). I agree with the sentiment here - it's more caring and passionate, to me still, to aggressively listen actively, interrupting and riffing off of one another whenever you can to build something beautiful together. If I have something important to say, why would I hold it in?

It's clear that we need higher-level system APIs. Is GTK actually worth it? What can we link with?

Subcultures are re-invented or recycled forever. "Cool" is no longer hard to acquire - it's who you know, what you wear, who you listen to, and what you consume - and all of these things can be purchased and commodified. This makes 'cool' more of an idealogical or informational economy than anything else - those arbiters of cool, those drivers of taste, keep knowledge exclusive and they're thought to be cool for it; they hold some status symbol that nobody can quite possess, because nobody but them has perfect information about the mission or end goal. This is influencer-driven fame, driving superficial copycats and subcultural churn as we move from one figure to the next.

The dream of this is some distribution of specification of subculture, where choices are not 'better' or 'more mainstream' - just different, and equally reasonable life decisions, based not on signaling but on lifestyle choices and the tastes they guide the user to.

Again - what does the AI model operate over? What machine-driven instructions can be crafted by humans in natural language in this way?

This speculative process kind of confuses me - surely these existing languages expose APIs very carefully crafted by humans, and as technologies and human needs evolve, we'll want to deliberately and consciously design and evolve the user interfaces to computing that we're capable of. No human can check that machine code will execute how they'll expect in all cases - and if we are to define some complete and inflexible specification for how a program should behave, this is no different from programming itself - albeit with a different paradigm.

Interesting - wondering what their perception of the technology or eventual vision is

This is a dream - to find a problem with something, make something on the fly to solve it, then refine it and make it available to others... the instructions, and making something easy to build and copy, are more important than being able to distribute a final product themselves. How powerful would it be to redistribute instructions and schematics for others to re-make in their own homes?

I swear a lot of these twitter jerk-refreshing bugs - jumping to the top of the feed, re-organizing or re-sorting it, etc - haven't been touched or fixed because they leverage this and keep you on the platform looking for what you just spotted

This is a mistake. Reusable, re-distributable components are vital to making changes work at scale. Elm introduces strange abstractions - CSS looks the same, but the DOM exposed by different parts of the page - though they may be intended to look similar - can be different. Betting against component-based architectures feels to me like betting against abstraction; I'm not sure if it's a winning battle. UI elements will inevitably require state to function, to add cool animations, etc., and there are ways to force this - but assuming encapsulated data in functional components allows for the most flexibility here, I think

Is reactive state the best way to go for a UI framework today? If Raph is doing it, it's likely one way to go; every time I've tried to do something computationally expensive with a framework like React, I've run into bottlenecks I've had to avoid by creating a canvas or some segregated part of the application wherein I mutated the DOM. Wondering whether we run into the same sorts of scaling problems with real-time audio/video applications here.

Not sure to what degree the "PhD" remark is - but this ratio of "stuff done to fix today" / "critical analysis of today" is quite terrible to be nowadays. Everyone wants to be the scholar and nobody wants to 'do the work', it seems. There just aren't enough industry levers to bring ideas to fruition.

Ironically, these quotes throughout the article to me represent the same kind of standardized, boring commodification that the article criticizes.

Take a break from appreciating a long-form article to share a prepackaged quote to Twitter? Seriously? I don't think the publishers understand that America's civic and digital monocultures are one and the same.

A ton of interesting video work is happening today - but theater and publishing monopolies eliminate 'traditional' movie theaters as a distribution option. Try YouTube, TikTok, Vimeo - you'll find beautiful, clever, and incredibly streamlined videos put together by individuals doing it better.

Why aren't they "establishment"? Unclear. Maybe some Disney execs need to get a few years older first.

[writing: needed some hyphenated characterizer here]

I don't understand the lack of personality here. Why is this reductive naming scheme a trend? "Katherine Small Gallery" is the best local counterexample that comes to mind - the place has a name and personality as quirky and elaborate as Michael himself.

How can we make the process of giving and receiving beautiful things to one another in a city as easy as the process of ordering furniture off of IKEA or Amazon?

I see this happen so frequently with those in local arts communities - as they're typically relatively low income and value beautiful, unique work - but not within cities at large.

It's also hard to justify some fixed cost down without guaranteeing some return on investment.. but with that fixed cost comes the ability to experience something beautiful.

I get this return through clothing mostly - I buy nice clothes [mostly used], then sell them at anywhere from 70 - 130% market, depending on what they're worth on the secondary market at the time - then put that money towards other new, beautiful clothing. The cost up-front and lack of trust here are both prohibiting though.

Who is designing these design standards and regulations? Why can't they be reconciled with diverse styles of building?

The attribution here is interesting - citing the prompt and version of Stable Diffusion that generated it should be sufficient to potentially identify what information it used to render the picture.

Beautiful. What can I make?

They've always been like this : )

Why? I don't see any evidence for this... company changes are incredibly chaotic but overall seem productive.

When reading pieces that express particular emotions that the author is experiencing, I love learning a bit about where they're at when writing - it adds more context to their emotional state at the time and helps the reader better emulate what they might be experiencing. I moved to the couch to read this : )

Source! I wish they supplied sources for this information. Where do I learn more?

Wondering what the literature says about this and whether it's actually good or valuable.

WHAT

Is this good?

A shooting game that includes arithmetic operations is just that... a shooting game with some shoehorned math chore to be memorized. The act of learning and exploring in and of itself has to be a core aspect of the learning tool or process or nothing will really be "learned".

Reminds me of Cornell's Bill Gates building lecture halls - but those had university seating rather than the pews of a church.

What happened?

Is the cause some association of computers with antisocial lifestyles? The curt and difficult personalities of many people who work with software? The reputation of "gamers"?

This feels like today's reality - we don't have computing agency!

Doing things is good actually

4'33" isn't a minimalist work because it contains so little deliberate sound. It is minimalist because it encourages the listener to focus on the ambient noise around them.

If we have control over the browser, it should be possible at the window management level to render some native text box instead of the textarea provided, injecting keystrokes and activating the properties necessary, but adding some modal editing functionality to the corresponding area.

I think it makes sense to bundle the browser with the window manager; the browser can't be given privileges, but it can be run with the same level of expressiveness. ChromeOS' web-first philosophy is one step towards this idea, but doesn't provide cool enough customizability or native tools to make it an OS for forever

Is it wrong for me to refuse to use this language because of this...

This is an oversimplification; the semantics are still concerned with a stack machine, and the user of the system has to hold the entire stack - this external system - in their context. The lambda calculus is only concerned with the current expression.

Brain is expressive query system that, given some constraints about some information, is able to retrieve lots of associated information from context. It's notably fairly difficult to demand precisely the information you want to receive.

A lot of the time some problem can be expressed as a query in some expressive database system more optimally

Why are there so many of them? It's okay to roll your own recursive descent parser. It's probably possible to abstract such a system, but not in such a way that uses another file format, specification or other, external system. Such a library should be native to the language it's expressed in.

optimized for usability!

Create a wrapper type for every measurement, then implement conversion functions to and from some measures A and B that untag and retag the values with types; units are implicitly enforced by constructors

What.............

...that adheres to GCC's "spec"?

tools don't matter; people do

isnt.online/~jake/

Interactive programming provides the same benefits - compiler errors are better studied, sure, but interactive programming provides very real benefits - they mirror the experimental reasoning of the real world! Unclear what's better here.

This is the context in which static typing is good, sure, but dynamic typing is far more agile when producing incomplete and buggy programs. Incomplete and buggy programs are wonderful because it's important to test programs in the real world and allow non-programmers to use them before they are complete; formalizing types is a lot of the work.

Answer: Estimate upper bound for stack space and work with this in mind. It's a "virtual stack" - in that we don't get the same benefits as using an L1 or L2 cache or something - but it's far faster than dynamically allocating memory as needed if we can do it all up front.

This is beautiful! I should try working in this way; can handwrite everything, then convert to SVG or similar for a scalable, compact web representation.

What was the impact of this program? Probably on the order of tens of thousands of engineers - right?

Why Legos in particular?

Consult multiple dictionaries. Learn more about the words you're using to explain yourself!

trust a dictionary. explore your words!

This is wild - requires long-term thinking and official credentials? There is some merit to these organizations.

;_;

super epic- believe in the vision!!!!!!!!!!!!!!!!!

Only barrier to entry to anything should be owning a smartphone

Learn more about these!

Type proofs are powerful - reasoning about progress and preservation motivates understanding of state and transformations.

Making work into play is creating immediate satisfaction. There is no reason for programming to be this miserable buildup of types and functions when we can run something, see it work, and build on it. Make more things feel like games!

This is why mathematicians say Lean is addictive or why people enjoy the REPL so much, or why tracking things more continuously matters. Smaller dopamine spikes closer together make things feel so much better and continuously motivate future results.

Yup. Making easy things works. Hard things may not.

Stream fusion optimizations are coming that eliminate a lot of constant factors, which are all that are relevant for most programming anyways.

Why don't we optimize for the lisp use case? We don't have enough tools to help these programmers working by themselves expand into other worlds. Ideas include: - Function stores with searchabililty, autocomplete, etc. to augment the mind while typing - Contract systems to make it easy to hand off some structured, typed interface to users of the code from other languages while allowing infinite iteration on the coder's side of the fence - Stronger interactive REPL tools and image-based development practices to satisfy those who thrive with immediate feedback

Lisp, by contrast, is just re-rolling everything

Code reusability and standards; everyone rolls their own libraries because everyone thinks in a different way, so there is no library continuity, but this is okay because the core standards are so expressive that being able to generate effectively is far more important than anything else.

After taking courses with John McCarthy himself?

Forth feels like a language that decays in usefulness in age; the biggest semantic strength of the stack machine paradigm, to me, is the ability to understand exactly what's in your program - and in memory - at any time. As the power of computing improves, there is less and less reason to care.

This is why! Insane.

I can't imagine what this looked like. Briefcase of punch cards that could be recorded and reordered?

Everyone should be able to cultivate their own garden of functions and carry them

What''s "real computing"? Lisp? Spaceships?

I can't imagine what this is like. Our feedback loops are so fast that it's insane to me that we're using the same tools. As Jack Rusher said: we should be leveraging machine feedback, not using "punchcard languages"!

So cool. A dream to live in that era and be in such a position at the time.

Most people didn't have these opportunities though.

: ')

I remember initially reading this before following the same path myself. Oops! I'll do better now that I feel so much shame.

Novel!

Thinking systematically and with structure is programming; formalizing thoughts is no different from preparing them for a machine.

Looking back, most of it's been thrown away. But it's important for prototypes to exist - we learn from them (sometimes).

This is an interesting way of evaluating a discipline: considering how many items must be remembered and recalled in active thought during the performance of a task. Machines can only get so far; holding the whole program in your head is the best way to go.

In 1972! How are we doing now? Techno-optimism is so, so important.

Reverse compiler to reconstruct the compiler again? Brilliant!

I admire the writing here: it's cute, but straightforward and relatively un-embellished - a bit like simple English - to communicate with the most broad audience possible. The videos! The translations! The "mocking" notes in the margins! The little touches of accessibility are just as important as the messages conveyed throughout the passage.

I think this payoff makes sense in large teams with verified software. If I'm at home, though, and want to get from 0 to working product as fast as possible, I want a simple tool that just works.

The difference between an ergonomic system and a stubborn, difficult one is how easily it is to write super readable, super elegant, super inefficient code in the system: the idea being that the readability of the entire system can be honed while the system is developed, while the optimization of the system can be honed when the system is maintained, after test cases have been accumulated and we can evaluate correctness for an optimized system by using the same interface as the prototype system.

this is so based

Whoever said this needs to rewrite Unix in purgatory

they were saying this in 1991 that's what's up

Parallel computing died when javascript only had singlethreaded semantics; webworkers aren't enough. it lets us do beautiful thing with ml models and software on someone's server, but we're definitely not leveraging them in userspace on personal computers in any meaningful way.

Why not tell the user to fix their code rather than getting it out of the way with the compiler!

We can do both...

Interactive apps are good: why would I want something to be text when I can interact with it?

Is this ironic?

Code is proof of itself; the ability to write code to compute some solution is proof by construction of a solution, and its behavior can be observed by demonstrating the code.

From what I've read about the process of Feynman and others, this isn't entirely true. Though there is power in being able to hold all of math in your head, the work is also on paper, or in the theorem prover, a symbiotic relationship that can track much more than the conscious mind.

Consider a case in which person A writes code that breaks the entire app but is on the way to some solution and they'd like to back it up digitally. Person B wants to receive updates from C's work on master, but they don't want A's to crash their thing and have to deal with all of A's garbage in a way that also respects their feature.

What gives? Turns out having that branch is good.

spaced repetition, if you deal in ideas

Critique, and offer suggestions, but not solutions

I like the framing of the article I read just prior to this - that productive engineering is making all of the correct tool decisions before approaching the problem, multiplying productivity by combining the net benefit of each tool.

Making these assessments and choosing to use these technologies is difficult. In particular, I disagree with the use of "architects" to administer some large-scale map of software that must be used to box in the product. I don't believe that these architects have the technical experience using the products to understand the impact that they have, and I don't believe that you can abstract this away as a series of bullet points on a slide - the ergonomics of a system and design patterns a system permits are difficult or impossible to express to someone who has no knowledge of software engineering paradigms.

I don't like the "do what you enjoy" mindset because I can find many things enjoyable; what's enjoyable doesn't tell me what's important, what matters, what I should care about or what I should work on. Make sure you're receiving new information from the right places.

Example: Hacker News is not a good place to receive new information. You will be swayed by startup juice and "cool tech" rather than ideas that matter to real people that exist in the real world.

The folllowing paragraph is brilliant! Contributes demonstrably to the message of the piece

I don't think that this is good - I love that programming is expressive and indicative of a style of a programmer! It's absolutely not the case that different programming styles are necessarily better or worse - obviously there are good and bad design decisions, but some tradeoffs are just different.

I'm not sure I completely buy this anymore... consider low-level programming. If I'm structure packing and writing to optimize, I don't expect most of this information to actually matter to the reader of the code, aside from the function names.

Does it make sense to give up on sight and start training with a synth or one of those word passing things on the screen? This idea that ideas can be absorbed and communicated faster than reading on paper or in conversation is fascinating to me - seems like a superpower.

Again, unclear how well seniority correlates with contribution!

Of course - because it's impossible to really identify a "10x engineer", nobody's going to throw you 10x the salary; the whole market will hedge bets and pay under market.

Bigger picture - it doesn't matter. It's not worth a workplace dispute or a re-evaluation. The code runs and it looks maintainable.

Yup... the prospect of planning a "career path" is kind of absurd to me. I don't want to play some game for promotions. I want to ship lots of things that help real people that sometimes correlate with adding value to the business. I don't think it's naive to say that focusing on making lots of real impact will make this happen.

Why did [last company] need so much orchestration for so few developers? Was this to mediate with contractors? We totally could have "hacked things together" without more expressive planning.

PMs/POs are supposed to mediate externally, but at [last company] they just seemed to mediate conversations and expectations internally. It's important for people to talk to each other - and it's important for the person building the product to be able to speak directly to their end user ad-hoc!

This is the biggest problem I've seen with previous companies I've worked with: they'll identify some customer need, then plan out some product without looping in the customer at all, push out the solution, and claim that this solves their problem. No iteration is involved, and bug fix requests or new associated features are put off for months. It's incredibly demoralizing.

This is what makes the software development course I took at Northeastern (https://felleisen.org/matthias/4500-f21/) so valuable. Software development was treated almost as a creative discipline - every class had critiques, wherein the creators defended their work in front of a panel that critiqued the work and attempted to find flaws, in the spirit of improving the code - together making software better.

Is this still true today? I look at Java, or the React app, and I see all of these design patterns and boilerplate that are getting in the way and preventing me from quickly developing code. Most of this is shared between devices.

Don't assume. Doubt, then understand. Only when you have a formal understanding (Could I throw it in Lean?) do you have an interesting result. Wonder how formal methods can be applied to thinking in this way.

updating priors everyone does this lol

preconceptions bad

Hero worship isn't the best attitude - but it's hard to look at what's been accomplished and say that there wasn't a predominantly positive impact made. Products aside, we're poring tons of money into battery optimization and large-scale sustainable manufacturing that we weren't before - key components of a future that requires clean energy to sustain.

(Others would argue that we need to give up on this world and reduce electronic consumption and waste - but this is an overly pessimistic (yet also overly optimistic?) view. Our lives are unsustainable in part as a result of consumption but innovation may be able to change this).

This is interesting in two ways! The will to question everything and the perception based on some observations that are fairly clear.

Insane! Toss the instruction manuals. Let kids play.

Computing is natural bottom-up. Top-down programming, by contrast, is an entirely different skill; poking around existing abstractions and developments to extract some relevant glue.

But it's clear that no matter what language you're using, you are programming with an abstraction - the computer is running your program, after all - so the framing of this writing isn't really the problem. The abstraction that the user interacts with just has to be better - it has to feel natural, like a part of the language, and not accomplish too much or too little so as to distract from the point. The building blocks should provide you help without getting in your way.

I'm reminded of Oglivy's perspective on advertising (who pioneered data-driven advertising techniques, and whose opinion very much concurs with this): advertisements are plain and designed to sell something. There is no secret trick - your brain is just satisfied in a simpler fashion than you think it is.

When folk rumors are echoed, people start to believe them - even if they're completely baseless!

Do not end up here.

Has the author written code before? Of course lambda is good. (Not as much in python, because multiline lambdas aren't allowed... but I wonder if the author has ever used modern javascript).

Both in some ways, but only Woz has enough historical context to know

This writing is just as beautiful and elegant as the message. The inner child matters. Live in a world of your own making.

Track my time with better granularity!

It sounds as if, code reusability be damned, there is room for some compositor solution that refuses to interoperate with other programs but "just works" as a complete system in a single compositor ecosystem.

I assume that this is what systems like GNOME try to do, but they compromise when trying to allow GTK to work anywhere; I'm perfectly okay with hardcoding all the programs I use at the WM level if it gets us these performance benefits.

Pro of coupling to Wayland: can use expressive free WM animation! Con: Nobody will ever be able to try your system on Windows or MacOS

mess is the process of thinking trying to overprescribe order to something that doesn't yet have a conclusive order is useless it also says something about prioritization: some things are more important tthan others, and a mess left means that we extracted what's useful

Oxide exists to build this good on-prem stuff!

This is bleeding into non-tech companies now, too... It's insane! Saw a video of the Staples logo re-unveiling from last week. Why are we getting rid of the cool L and replacing it with a somewhat curved U shape? You can't tell what it is!

It's important for design to be cohesive, sure - but it's worse, IMO, to have slightly clashing designs than radically clashing ones. At least with the latter I can tell the difference between my apps.

Takes a few years

;_;

How do we know that they outperform? + Surely they can be easily manipulated if relied on? I don't understand how letting more people throw money behind ideas is important - any financial prediction market allows for voting proportional to wealth, which skews all opinions towards those who own more.

I am definitely doing this too... this is probably a bad idea. I should put in the extra effort to polish things and make them accessible to others - something the Clojure community does right - and this starts with some sort of type system or other boundary.

Do not

I hate that I relate to this. I need to fix my work!

Leverage holes and difficult processes! Make them easy and elegant and seamless. Automate human effort away so people do things that only people can do and that people can do best.

just has to meet us where our abstractions are; doesn't have to ground to reality

Imagine having this level of mind-share

I know this is a bit of a snarky, offhand comment, but I love this deference of responsibility - saying "I have no desire to learn about this thing" is powerful

Again with the quips - and a break of the fourth wall! Brilliant outlook.

Private blockchains are useful insofar as they are linked lists with data that can be exported as a transaction log; that's all. You trust everyone involved via traditional business relationships.

So true.

Also - love that the author knows how to bring definitions in without a second thought, knows when to interject with casual thoughts to make the reader feel more comfortable. This piece reads like a smooth, continuous river of knowledge; definitions enter your mind seamlessly and are brought with you as you continue to read the piece, and the right words comfort you - meet you with the correct level of technical knowledge and casual attitudes - is a wonderful approach to writing. This is, honestly, the best "pseudo-technical" document I've read - it's a deep dive, not pop science, but it's accessible enough to communicate complexities that need to be understood.

don't try to make this work ;_; bad idea in most cases, unless transaction occurs in eth

A precursor to Lisp. What happened?

Astute observation. We still don't do this well today! We need better ways of inputting patterns; currently we have drawing and writing text, but how can we consider this process of thinking of information to fill in the gap in a more interesting way?

Lean's "library search" tool, most recently, blew me away with its ability to leverage all of lean's mathlib to fill in proof holes for free, replacing what would be an incredibly painful procedure of redundant, clerical work with a single computer command that summons the power of all of mathematics to finish your proof for you. Insane. What happens if we add synthesis?

The point of technology is to reduce this waste of time: if we have some gap in knowledge, we should be able to fill it immediately, by conjuring up information - not just some nominal text, but a structured data representation or visualization that helps me discern information from that text. Innovations in computation to help the brain think all involve reducing the barriers to these thinking tools as much as possible.

What is this? Look into it.

Interesting! Licklider here reminds me of Patrick Collison's thoughts on fusing natural language systems as boundaries between companies and similar things. There is a world in which we leverage artificial intelligence to continue to automate boundaries between people and machine as boundaries between machines - removing the work people need to do to communicate. Most process improvements today, I think, involve replacing human communicative "glue" with machines talking to one another; removing a human from the loop in so many (of the correct) processes will drive success.

By "correct" process, I mean that secretarial or clerical jobs should be eliminated - they're just humans performing these pure data transformations that software is already great at. Causes that require moral reasoning, such as lawsuits and principle development, should never be automated or left to computers. We should all be able to do the latter!

In other words - we don't need to use the classroom to fight to escape uniformity anymore. Social media lets us express ourselves as wide as we can to whoever we want, however we want, so providing an additional outlet for self-expression doesn't feel any more agentic than crafting an instagram feed. Rather, social media keeps us from navigating our local communities - and this is a practice that can stimulate friendship and growth, inside the classroom and out, to motivate students.

The other things that this article touches on are handled well, but I haven't seen any good projections or mentions provided for productivity graphs. This is far too hard to measure; the best measure is found in the people who use the product.

I really appreciate this approach to software development - in stark contrast to accumulating debt over the course of a large software project. If we had a software engineering degree, this should replace OOD or something - some "rapid development" course.

aesthetic of performance, of efficiency, of effectiveness. it's different than the senseless luxury and social conventions that people expect otherwise. big tech is tilting towards the latter. i really don't care for free food, but give me a building that i can route myself effectively through and surroundings i can enjoy interacting with

I have the former mindset... it's okay to have a terrible apartment if you can afford to be out all the time!

I love this way of working. If you're effectively operating like an individual consultant, you're held directly accountable for exactly what you accomplish.

Direct accountability isn't practiced as much in software anymore because people don't understand the consequences of making changes to the systems they're working on. This is stupid. Creating accountability is synonymous with codifying expected behavior and assessing deviance from it. It's good to test, actually.

;_;

Negative reinforcement with deadlines doesn't work - it doesn't push or provide enough opportunities to learn more. It's safe to assume that students have a natural compulsion to learn, but they're also presented with tempting opportunities; how do you provide a stimulating positive reinforcement cycle that sets students on the right path?

In other words: it's unclear whether ceding management positions to MBAs is relevant because real technical decisions have to be mad,e but the alternative is losing the productivity of an extremely effective engineer to do so.

This is wild! Modern technology radically changes the way the Lakota live, independent of any sort of colonization/

I am so sorry. Is this useful?

Aren't both of these ways of conveying opinions too? : )

This is cool! I want to be able to mimic and record and reproduce all of my movements and work.

Why? Is there some interest in improving the way they interact with their work on a physical level [like improving the ergonomics of sending an email or joining a virtual meeting or browsing the web]? How do we measure different types of thinking?

Progress in bridging the gap between programming a computer and using a computer. If we offer higher level customization facilities, people will be able to learn how to use these tools to accomplish gooooood things! Such as making their computers fit them.

I wonder if this kind of social pressure is overall productive. Driving smart people to elite institutions rather than pursuing (maybe arbitrary) personal passions might be sort of "globally" useful for progress - but it does build the prestige of these institutions, centralizing knowledge and ability, which is a questionable effort.

It's important to build an interface with semantics that make sense in any medium. Can we network it? Can we share it with a friend? Can we write it on paper? Can we understand with names alone? If these are true, it's probably a good library.

Everyone deserves the right to share. Simply and beautifully put.

I adore Stallman's writing style. Sentences are brief, uncomplicated, and straightforward. He uses very simple English except when he specifically defines terms and uses the terms he defines, so every sentence has a single, unambiguous meaning. The writing isn't entertaining in and of itself, but that's okay; the principles are quite simple, and from his strict principles come a wealth of wonderful ideas.

So you can just pass pointers? Insane. Brilliant!

Holy shit what

Absolutely insane

That's what someone who isn't would say : 0

The self-identifiation with this "rationalist" community and a corresponding traditional mindset is a bit confusing to me - to what end is saying "i am a rationalist" as a series of views part of their identities? Is this an abbreviation for some broad spectrum of views, or does it mean something slightly different to each person?

More importantly, how and why does it prompt all of these dating posts on and pages on the internet?

Jalex is a character. I'm not sure what to think of them, but I like the layout of their website. What's a theorist?

How can we re-architect our approach to primary schooling in the real world to bring these talents out? Social skill development is by definition punitive: you're hurt for not knowing rather than encouraged to develop. The internet gives people such incredible spaces to thrive, but still shuts people out of the real world!

Insane but totally possible that a human is born with a brain that functions exactly like the US legal system

School should be subverted as a tool to reference while practicing; it's a shame that it's not. Taking classes after working in industry gives me both an immense appreciation for the work I do in class, and better informs me of what is most important to spend my time on while I'm still a student.

I love how this "stanza" is written; the narrative of progression here is incredibly clear.

Is this good? Noble? Should I livestream myself working and living like this?

I'm not sure I like that the experience is, to a degree, performative and uncurated. I like having the ability to ruminate and collect information before broadcasting it.

I would love to be a wikipedia contributor, but I'm honestly not sure how or where to start!

One other thing I've noticed: the more active I am on the internet, the less active I tend to be in real life, and consequently the less happy I am (for some small happiness delta; this is a portion of the 'social satisfaction' happiness category, not the whole pie).

People like Knapp are seriously heroes. There is no tool better than knowledge and no role nobler than spreading it. I have nothing but admiration with his societal position.

is this insane?

This is very high level, but it's absolutely feasible; Mastodon is a technology that exists, and using a different recommendation system is not unreasonable. This says more about the knowledge of the author of this article than of the texts.

I hope you do, Grady : )