technology ouroboros

Why does computer tech always seems to reinvent itself eventually? Example 1: in the 90s they created HTML for static websites, then they figured a way to render a page on the server using php in response to user actions. Then they invented javascript to make client-side interactions possible. Then they over time transferred page rendering from backend to frontend. Then they figured out that client side rendering is slow and created soykaf like server-side components basically reinventing php but slower. Example 2: lisp was a king of everything non-unix, but 8 and 16 bit computers couldn't run it properly so it was forgotten in favour of C. C wasn't portable enough between architectures at the time so they invented Java. Later java's verbosity started tk really annoy people so they started inventing languages "java but better", eventually creating clojure which is java lisp. Lisp got reinvented again. You could probably find a lot more exampels too. Why does this always happen in computer technology and pretty much nowhere else? I know other fields sometimes take failed ideas and make them work with new tech like stealth bombers but it's different from independantly evolving into reinventing existing older tech.

1. Pioneers develop stable and usable tech 1. Grifters take over their tech and remove the pioneers from the steering wheel 2. They replace good workers with average workers (JS era), then below average workers (AI slop era) 3. Average workers are midwits, they are easily impressed by shiny toys. They are oversocialized creatures, they form insider trades where they propagate their mental illness to the masses. They readily destroy their own foundation previously laid out by the pioneers. They create nodejs, electron, npm packages like is-odd is-even. They jerk off their "Uncle Bob" and "clean code". This allows grifters to wash more money by forcing companies to switch their tech stacks to the unstable mess. Unstable mess requires constant maintenance = more money for grifters. The smell of a rotting corpse of the tech field attracts more parasites and vultures. 4. Meanwhile average workers undermine and destroy themselves, allowing themselves to be replaced with below average workers. 5. Computers went wrong when you made them for niggеrs.

>>93785 first of all, your timeline is completely wrong, and so are the motivations that you imagine were behind those technologies. there is not a single thing in your entire premise that is not incorrect. and second of all, it seems all the same to you because you only have a surface understanding for these things >>93788 that's a fake quote, and even if it was real, linus loves ai contributions, he's doing the thing you call below-average and your narrative is a stupid myth from *chan. most "pioneers" were bad, like, their code was both slow and unmaintainable, of course, you have never read it, so you wouldn't know, that's why these narratives spread. like, just look up computer science papers from back then, you will see that what was considered "expertise" back then is basic stuff now there doesn't need to be a glorious past for you to not like javascript. the modern hobbyist has a better computer science background and more practice programming than most of those guys had when they were making soykaf in the 70s. these tools are widely used because that's what was available at the time, you don't need to turn them into your father/ancestor figures

>>93792 >their code was both slow and unmaintainable What exactly about "develop" part in "develop stable and usable tech" you don't understand? How do you think development works, you bloody monkey? These people went through all the painful experience while you still were a sperm floating in the nutsack, so YOU don't have to. Discovery works from the ground up, you make mistakes, you learn on them, then boom - you have a foundation of experience. All of that just to be disrespected and discarded by soykaf ters like yourself who think they are smarter than anyone else, yet all you do is read moral lectures to strangers on dead imageboards. So much for being a smart guy. The modern hobbyist has nothing to do with your average tech worker who barely knows any bash, you arrogant clown.

>>93794 nta but I don't blame them for not knowing the poopy shell language. POSIX shells were widely disliked in some circles (not just lisp hackers, some unix graybeards also preferred Csh over poxis soykaf) Although, I agree with basic premise. The anon basically says that Newtone was a retard because he sepnt so much time studying physics than is now taught in middle school.

>Why does this always happen in computer technology and pretty much nowhere else? Ever heard of the steam engine? We just find more convenient and sometimes more efficient ways of doing the same thing. We iterate and iterate on the same familiar concepts because sometimes you don't need to reinvent the entire wheel.

>>93797 The thing is, in computers the reinvented technology is usually straight up worse that what was before. It was documented long ago as Greenspun's 10th rule.

>>93800 that's the software developer mentality, they wanna do something new and cool to completely flip the script and change the paradigm

>>93794 so that's your answer? that yes, their stuff was garbage compared to the modern alternatives, but that I'm wrong because I'm being mean to them? you're a racist piece of soykaf, and not very mature judging by how much you have let imageboards influence you (so far in this conversation you have expressed zero original thoughts, I could train a llm on g and it would be indistinguishable from you). so why do you act all offended that I'm not being civil >develop stable and usable tech they wrote soykafty code, "stable" doesn't mean anything, specially back then, and "usable" just means that you can use it, which is an extremely low bar. people have been both trying to improve it, and to discard it and start again but with their mistakes in mind, and you complain that, because the result is still bad they are midwits (even though their stuff is way better than what the "old masters" did) >>93795 reading comprehension, what I'm saying is that modern physicists know more stuff than newton did and that it's stupid to say >oh all these modern guys are so dumb compared to newton furthermore, it's a bad analogy because newton worked on more theoretic grounds. writing something fast and maintainable is part science (data structures and such) and part craft (actual code), and it is in that part that they more often failed. you can call their code "primitive" if you want but that's just an euphemism, it's bad

>>93802 >you're a racist piece of soykaf What's the matter, reddit boy? Frustrated there's no downvote button? Awww shucks my sweet little summer child, here, take my upvote as a sign of armistice. :3 >modern physicists know more stuff than newton did You are so on point, my mature queen, they know all that stuff that was already written to books thanks to guys who actually came up with that stuff. Like, they know strings theory (they couldn't prove) or bing bang bong theory they couldn't prove either, or bubbles theory they couldn't prove, and erm multiverse hypothesis with zero evidence (but at least it sounds cool) and basically there's been a stagnation for 30 years but at least my little black rocket scientists are educated and knowledgeable on... *checks notes* stuff, yeah, that's right. >to discard it and start again but with their mistakes in mind When my black queen speaks, I sit down and listen. You are so true sis, and this is exactly why they reintroduced all the old mistakes but in a much worse, slower and bloated way! The myth that rich and powerful guys deliberately ruined the tech field with cheap workforce and made coding "available to cretins" is so stupid. Like, entirely unsubstantial, we all know everybody was working hard to improve coding, like doing actually important things, like installing is-odd and renaming master to main, like. Soooo true sister, that's like entirely right-wing and conspiratory, and we all know conspiracies are dumb and stupid. Punch the nazis!

>>93788 >whole two packages to check whether a number is odd or even actually exist and are serious "projects" what the actual fuarrrk geg the first thing that comes to mind for me would be to check the least significant bit because that's literally just it it's so absolutely confounding to know that there are people that actually do soykaf like that

>>93802 Maybe if you had a little bit less bile in you, maybe people would have liked you better. I just refuse to engage with someone who speaks in insults and hubris.

>>93810 For my whole life I've been doing it like this: > x % 2 === 0 > x % 2 === 1 Anyway that whole thing makes sense to me: > whole two packages to check whether a number is odd or even It comes from an era when people wanted to push this idea of modularity to extreme, to the extent that any single helper function could be in its own package. It's elegant/beautiful in a way, but those dreams got crushed by malicious actors compromising npm and pushing straight up malware under similar names. And those security/ownership concerns are the only real downsides. With native languages there is a concept of compilation unit, so it has additional downsides regarding compile times and optimizations I guess, but this is not a concern for JS.

>>93810 Best part is that is-even depends on is-odd and just does `return !is-odd(n)` The reason is-even actually exist though is because it checks if the number is a number and is an integer. The number check uses is-number package. JS solution to bad tyep system is making 1000 packages that do boilerplate code.

>>93810 used by a grand total of 0 big projects - all the npm dependents are hello world projects and/or empty repos with a readme. you can also make a library like that in C and then pretend that C programmers are incompetent because there exist a libisodd. it's such a dumb point that I'm losing brain cells just from writing a reply >>93811 baby wants to cry? see my reply to the other guy. it doesn't matter if you engage or not, because you have no critical sense whatsoever, you just accept whatever lie is convenient for your narrative without even checking

>>93813 > JS solution to bad tyep system is making 1000 packages that do boilerplate code. It's not really that, packages like lodash, underscore, ramda, etc, existed since forever. It was genuinely an attempt to push modularity to the maximum.

>>93812 >doing modulo to check even vs odd i mean fair enough if the compiler optimizes it out (strange case but it's plausible) but is le wrong and le bad because unnecessary extra cycles because division and that >modularity i think modularity of the program that is being developed is important, not being able to throw a billion different things together. when i do do things i try to write them in a way that all of the parts that actually do something significant are their own abstraction layer, that way it stops you making monolithic files and allows for easy rewriting of a component if the apis are clean. i think "extreme modularity" is probably a bad thing because everyone ends up relying on a heap of different libraries, makes things fragile, doing that as well probably (generalization) hurts your ability to understand how everything actually works. having libraries for things is great but i can't see myself using anything outside what i'm making unless it's something very large or something i cannot afford to mess up (e.g. cryptographic stuff, http) >>93815 idk i wasn't really saying anything for anyone, i just think that something so dumb existing to start with, was dumb though, as i said just before (or whatever parts are relevant) which i think was basically "having a billion things le bad because you don't understand it and don't learn" >hypothetical point against c programmers because <retarded thing> exists there too it's not really about any specific language, but it does make slightly more sense in c because there's much lower level control over what the machine does. < whatever i'm saying is probably incoherent but there might be some meaning it in

>>93818 you can upload whatever you want to npm, that's why it exists. no one actually uses it imo there is a point where modularity can hurt readability, like you say, it makes things hard to understand

>>93818 You seem completely out of touch regarding values and concerns of high level software dev. > i mean fair enough if the compiler optimizes it out (strange case but it's plausible) but is le wrong and le bad because unnecessary extra cycles because division and that This would be huge red flag at webdev or enterprise dev interview if you said something like that. Those values are niche and only applicable in low level systems programming.

>>93819 > no one uses it is-even has 693000 weekly downloads

>>93821 https://www.npmjs.com/browse/depended/is-even it's only used by 63 other npm packages, all of which are either empty, vibecoded, or some tutorial on how to use npm the downloads number is just bots >>>/tech/93820 >would be totally not a situation you made up in your mind to get angry at whatever you think "high level software dev" is

>>93820 why would i ever wanna work for Germans like that geg

>>93822 > https://www.npmjs.com/browse/depended/is-even > https://www.npmjs.com/package/kill-your-self Kek >>93822 >>93823 No but that's really like that. There are even slurs in some languages that high level programmers call those who "prematurely optimize" and all that. Values of low level hackers are often in direct conflict of interest with those of high level devs. Your job as high level dev is to assume everything is a black box and compose black boxes most productively, when you're trying to look inside those black boxes at every single step and especially in absolutely ridiculous cases like modulo division, that's exactly the most wrong mindset ever for this kind of work.

>>93825 ok, but it's not relevant to me because i'd never waste my time """working""" for a company that operates in such a way, since i have no interest in money, and i actually care about doing things in a reasonable way instead of making slop

>>93826 I'm personally not entirely sure all problems are approachable this way, but I also want to believe it's possible to keep things simple or at least decompose projects in a way that allows keeping separate pieces managed at reasonable level of in-depth understanding while still having them work together to achieve complex behaviors.

>>93800 >The thing is, in computers the reinvented technology is usually straight up worse that what was before. Ah, allow me to tell you about the steam engine then. You see, the steam engine which operates as an external combustion engine at the beginning of the 19th century had a competitor called a "Stirling Engine," invented in 1816. The Stirling Engine is a closed cycle engine which was notable for its dramatically increased efficiency thanks to its regenerator. A Stirling Engine running at proper high temperature has a thermal efficiency of 67%. This is the maximum efficiency achievable in a thermal-mechanical engine. Since a Stirling requires high temperature to achieve this efficiency, the steam turbine engine--dramatically less efficient, but capable of operation at lower boiler temperatures--was preferred. Boiler safety improved, and with those improvements, no adoption of Stirling engines and no scaling of Stirling designs ever occurred. The engine is now known primarily for its role in prototyping designs for reflective solar collectors. You see, in addition to the material difficulty of maintaining correct temperatures in power generation, it was also found difficult to maintain teams of engine designers who could design and specify high efficiency Stirling engines or to build Stirling engine components instead of modifying teams of Steam Turbine engine engineers and designers and machine shops to precision-machine Steam Turbine components. Ah, but, of course, there are now entirely new electrical power plant types that did not exist in the 19th century, why Nuclear Power is of course something that was redesigned from the ground up, aha checkmate atheists, yes? Well. Nuclear power has a kind of failure mode that even coal power does not have. You see, the nuclear power system is actually put into a great deal of jeopardy when it runs too hot. Components have a chance of melting, and, you see, when some particular components get so hot that they melt in a downwards direction, well, then they make a sixteen part documentary series about you and your life choices. And that's not a life goal for an aspiring steam engineer. So, you engineer your atomic power source to run even cooler than steam locomotives of the 19th century did. So the high-efficiency steam turbines of modern steam turbine coal power plants are utterly useless to you, let alone high efficiency Stirlings. A steam turbine for a coal plant runs on dry steam, well that's something of a misnomer really your turbine is technically running on something closer to a plasma let alone to molecular steam. A turbine for a nuclear power plant in contrast must run on wet steam. It must be vastly larger in diameter, vastly slower, chemically resilient against the condensate of both water and your water additives, and in many if not most objective terms it performs significantly worse at much higher cost than a coal turbine. Fusion plant turbines promise to run at worse efficiencies due to even worse conditions, but we'll know the real world machinery when we get those real world machines. We will never, ever, ever run our power grid on anything that can match the 67% coalfired Stirling.

>>93811 NTA but you are way off the mark. Let's review the chain. >>93792 makes his point to no one in particular. >>93794 comes out swinging with derision, "you bloody monkey", "soykaf ters" and "arrogant clown". Then he writes the post you say has "bile" in it. When you talk in earnest and get people just insulting you, it's hard not to snap back. I try not to, in part because I have a genuine desire to talk about a subject than just insult people, but also because a reasonable person (ideal target audience) will recognize the situation as the battle of wits of the idiot and academic. I don't even think the guy made his point well enough, but he comes off better compared to that geriatric or suckup or whatever he is talking up the old school guys. To be fair, I think the old guys figured out stuff a lot sooner than we give them credit for, plenty of actual problems were settled by the 60s and development after that has really been in the way of plain standardization. For instance, converging on the 8-bit byte, IEEE-754, converging on Unix and later Linux, etc etc.

>>93822 As a reviewer who claims he was talking "in earnest", you need to get your eyes checked and a pair of glasses. Somehow you "reviewed" the chain and missed the entire reason why that guy deserved the derision. Hmmm, why would that be? Let's think thoroughly: a) maybe because he replied to OP with condescension, claimed he is incorrect and has "surface understanding of things" (mocking him) while providing no counter-arguments whatsoever? Just throwing a turd into the void. b) maybe because he does not understand humor and has a need to point out a fake Linus quote, posted on an imageboard out of all places? Maybe because he does not recognize neither image (which is quite old) nor the T. Davis quote (he spergs about muh racism later), which points a huge arrow at him written "I AM A TOURIST." c) maybe because he called it a "myth", then claimed the pioneers were bad, then claimed "you have never read it so you wouldn't know" (blatant projection and attempt at ad hominem), then claimed what they discovered on their own is "basic stuff" now, so modern guys are "better" somehow. So basically he did call Newton stupid. In his second post he tried to deflect and deny he ever did, pointing out "reading comprehension", yet reading comprehension clearly states that he did. He is denying the reality and is unable to deal with the consequences of his own words - suggests low impulse control and/or low effort put into the thought. d) maybe because he claims modern developers are better, yet all we see is software getting progressively worse while shipping more spyware and hogging more cpu and ram, and bringing absolutely nothing new to the table and nothing of substance to the end-user? My own pair of eyes right now says that electron applications hog about 2GB of RAM on average. What new features do I get? I don't get any, but it makes developer's life easier - at my own expense! e) he claims "stable and usable is a low bar", yet modern software proves that most tech workers are behind that low bar - especially the ones imported from India. Not a racist remark, but if you believe they have the same compsci background as your average white or black american guy, you are an indian or an idiot. f) maybe because he's a tourist from reddit and is trying to mock *chan culture while posting on a *chan himself? g) maybe because he lacks intellect to understand the chain of causation and does not comprehend that you first need to write soykaf software to make something better, or in other words, you have to fall from a bicycle to learn riding on a bicycle. He claims modern hobbyists have a better background, yet ignores why they have it. Because the "bad pioneers" have built that exact background - pointing that out just in case of particular density. Yet tech workers, as explained earlier, do not have or care about having that same background, so it literally does not matter. Corporate software world is all about stepping on the same rakes - which is something you would know about working for a corporation yourself, and this answers OP's question directly as to why they constantly reinvent the same stuff over and over again. So, why would that guy deserve the derision? Gee, I don't even know. Oh, but you are a smart guy and can figure it out yourself. Clearly you are not just sympathizing with a reddit sperg tourist.

>>93834 you're not melting, are you and nta, but I wrote the post that you're whining about and: >a there are no arguments to counter, no substantiation or evidence. OP (you) has a surface understanding, that's why his post is just over-generalizations and vibes: memes he picked up from imageboards and uncritically accepted. you don't need a debate to see this >b if it was a joke you wouldn't be so pissed. you said something stupid, I disagreed, and it hurt your feelings (which I did not intend to do), now you're claiming that you are merely __pretending__ to be retarded. but at least you're aware enough to admit you have the inner life of a gpu, no creativity, and no identity or thoughts besides stale soykaf you picked up from -chan >c severe reading comprehension issues. you want to sound reasonable but you're too stupid to understand that to do that you have to make an honest attempt to understand and portray what the opposing side is saying >d except the pioneers were making kernels, compilers, CLIs, not websites, it only makes sense to compare them to the people making kernels and compilers now. the anonymous people making internal corporate software back then were just as checked out, the only difference is that they had worse tools and a worse theoretical background. besides, before you were saying that the modern guys are incompetent, now you say that they are just pursuing different objectives >e things like kernel crashes are extremely rare now compared to the 70s, use your brain for a second >g so you admit that the "old masters" wrote bad code, that they knew less than what is known now, but are too infantile to say it openly because... that would hurt the image you have of them as perfect geniuses way above modern midwits. but we are getting to the root of the problem: you think that saying "X didn't know Y" is the same as calling X stupid, which is why you can't admit you don't know soykaf and act offended whenever I point out you haven't even read the wikipedia article for any of these technologies

>>93839 As an OP I obligated to say that I did not participate in your discussion except fot calling out unnecessary insults that clearly don't steer the conversation in a productive direction. To which you replied cry about and I ignored all the subsequent text walls because I can read /g/ if I want to read insults about technology. Really, the threads premise is a giant stretch the second example doesn't even agree with it. I do understand that there are genuine uses for react as someone who once wrote and ad-hoc framework for generating a page (it was a lot of document.createElement and pain). I also do know that almost everyone is aware of server side website generation because any backend course for noobz will pick something like express.js/flask with a template engine. The thread is basically about using overcomplicated technology where it is for the most part not needed or even more basically it's another "bloat bad. I want my html internet with minimal javascript back" thread. As you can see, I know my soykaf, but I just tried to pull something out of my ass to make it look more interestin and in the process maybe find out if we have any other curious cases of circling back to doing a simple thing in a complicated way.

that guy is just too retarded to argue with if he is unable to understand the essence of the >>93788 which basically said "caring people with lots of knowledge got pushed out and replaced by uncaring people with lots of money who wanted more money" which is a cemented fact known to anyone who was present in the tech field between 2011 and today, and to deny that is to prove the assumption of severe retardation. so again no point in arguing with that guy. waste of time fighting against their ego instead of fighting over truth

>>93846 >"caring people with lots of knowledge got pushed out and replaced by uncaring people with lots of money who wanted more money" Come on lad, anyone could make a game of explaining away the paradox of progress in any field with reflexive simplistic blackpilling. "Why did Stirling engines with a 67% efficiency lose to coal steam turbines with a 45% thermal efficiency cap, which in turn are increasingly replaced by nuclear steam turbines with a 33% thermal efficiency? Moneypeople are the boogieman, nothing ever gets better." The paradox of progress didn't just hit businesses and the hobbyists who fetishize entering into said business. Greed is no answer. The young cabal of informed and knowledgable ambitious young men who wish to reform tech according to bold theoretical improvement and are tragically pushed out by a callous establishment is currently embodied by the movement to rewrite C in RUST, with the rustaceans as the knowledgeable people who care far, far, FAR too much. Incorporating Rust into the Linux Kernel, the single most controversial Rustacean move, did not and would never involve any attempt at relicensing the kernel. Where does greed enter that controversy? You, personally, aren't a LISP hacker. You read SICP, learned LISP in some Scheme form or another, lamented the fall of LISP machines, and then after some point just never touched LISP again. There is no body of "caring people with lots of knowledge" who got pushed out of high profile positions but could have done The Coolest And Best Thing, and some scrappy underdog caring person with lots of knowledge is not going to Do The Thing That Everyone Overlooks. Nobody's going to make a 60% efficiency carnot-stirling engine motor vehicle, nobody's going to make a LISP OS that replaces Linux much less Apple or Windows, and web standards are the mess that they are because web standards have to serve web users and web users enjoy breaking web sites.

>>93847 NTA, You're comparing a field where large corporations competed very hard on R&D for almost a century to get the current state-of-the-art, to one where large corporations mostly rely on rent-seeking and anti-competitive practices for the past thirty years to get the current state-of-the-art. Software manufacturers are remarkably slow to innovate, given the flexibility of the product. >You, personally, aren't a LISP hacker. You read SICP, learned LISP in some Scheme form or another, lamented the fall of LISP machines, and then after some point just never touched LISP again. You don't know that. >web standards are the mess that they are because web standards have to serve web users and web users enjoy breaking web sites. It's more that they were implemented in a very ad-hoc fashion and people mostly went with whatever was the fastest to ship, and then immediately fossilized it. >nobody's going to make a LISP OS that replaces Linux much less Apple or Windows Because it would be impossible to compete, or because nobody would invest enough time and money to make it happen?