4:21 I think the "Lindy effect" is really powerful in this regard. Paradoxically, it's often the oldest stuff that is worth learning, that will stick around for longer than whatever brand new shiny thing has appeared at the cutting edge. If it's an old book and people are still talking about it, it's probably worth reading. It's not likely that someone is going to come along and completely shatter our foundational understanding of philosophy and mathematics. Not impossible, just unlikely.
@@EdwinMartin The Lindy effect is just an observation that whatever has survived a longer time is more probable to survive in the future relative to recently produced things. That doesn’t mean new stuff won’t replace old stuff.
@@EdwinMartin Arguably, no. Classical physics is still the foundation of physics in engineering. Newtonian mechanics is in no practical sense obsolete because of relativity or quantum mechanics.
1:13 I actually did take a long ass break from webdev a little after jQuery got popular, and coming home to see paint everywhere is a totally accurate description of what it felt like to find out about shit like Node and React.
I still don't "get" Node. Yes, the react pattern was cool. Yes, it was neat how you could start a web site with a tiny amount of code. Neither one was a Javascript--specific feature and Node is (IMHO) a bad implementation of these patterns. Which gives rise to the supposed "advantages" of using the same language on the client and server. "Advantages" that have never materialized in the data I've collected on productivity. Not even once.
@@thewiirocks All node is, is an extraction of google's Javascript engine outside of the browser, so you can run Javascript in other environments, server side, desktop application etc. Never made much sense to me as I was never a fan of Javascript. But it was never aimed at me, it was aimed at all those people that only knew Javascript.
I think that transpilation is an insane concept. We are picking interpreted languages and treating them like compiled languages. I'm glad that Rust exists now to fix this madness.
@@rumplstiltztinkersteinRust is on a rushing change also. But it has a strong typed system and a strong compiler. So everything is always kept on the rail of compatibility. That's really amazing.
13:45 I worked on a project for 4.5 years and rewrote it twice in that time, about 3 months each. The DB greatly changed in the first rewrite and was sharded in the second. Huge changes, other than the Excel library (Ignite UI) we used. Write your code like you intend to rewrite it every few years. I've lived by that since I first heard it and it works so well with the evolution of frameworks and architectural decisions.
It's indicative of slow pace & small size of your team. Anyone who has worked in a fast paced environment will confirm, that amount of code often so is large, that it's practically impossible to rewrite it. Not mentioned the political capital you need to pull that off. Additionally, in companies with high focus on personal performance (which is all top companies) such refactoring is detrimental to your career and can get you fired.
@@S-we2gp Mostly that I don't expect anything I write to be around forever. Technology changes, org needs change, life moves on from what you wrote. Even staying in the same place for 3 years, I rewrote everything once to improve performance. Maybe it's different for people not in Full Stack roles...
Building the first 80% from scratch, thats the really enjoyable part. Over the years, I've carefully positioned myself in my company to get in on a new project, build it up to almost maturity and then move on to the next interesting project. Working too much on the last 20% will lead to burnout, or at least that's what it feels like to me.
I made the decision to just switch to Rust and C++ for this very reason. Yes those languages are hard but there's a sense of completion. Also I couldn't care less about React or Next.js (It feels taboo to say that out loud)
Out of curiosity, what are you working with atm? I am also eager to escape web and Rust and C++ look very interesting to me but struggling to pick a field because I don't really want to do games.
@@thesaintseiya Solidity is my strongest language, even though I worked on JS / TS for longer. I feel you on the game dev path. That's another pro of low level languages. Career flexibility: Games, VR, Robotics, CLI tools you name it. Solidity is awesome BUT it is specific to blockchain and the Ethereum chain. If anything happens to Ethereum, you are screwed. Right now Arbitrum are working on a tool that will make it possible to write contracts in C++ and Rust... That should tell you something.
@@hyperthreaded I may look into it. But that's a HUGE maybe. Just bored of web. 🤷♂️ Also WASM is pretty niche as JS dominates the web. I'd rather use Rust and C++ to do things that JavaScript can't do.
3:40 We also have a lot of applications that are stuck on older package versions. Every time time a vulnerability gets discovered, we need to upgrade all applications, as it should be. But often enough the fix was applied in some next major version. Support for other than the latest version is often very poor, forcing you to upgrade. And since it is a major version upgrade, the solution is often non trivial. To keep up we need to have one persion that does nothing but constantly upgrading the applications we're responsible for. This is definitively a major downside of a micro-service/frontend architecture
Mullen is a legend. No exaggeration at all. It took decades for freestyle like what he helped pioneer really start to be accepted back into the mainstream of skating many years after the whole X games phenomena.
The best skill in web development: Focusing on your foundation first and ignoring the hype of the new tools. Frameworks are great but only after your foundation is strong.
I know how he feels. Working on my personal site is a lot more fun than working for a corporation. My current site is my second attempt at an SPA, and the first that doesn't break the Back and Forward buttons. I write all of my code from scratch with no frameworks, borrowing snippets from Stack Overflow as needed.
9:50 You know what Hasn't changed that much... Binary. The language literally has only two functions. 1 = true/on and 0 = false/off. I mean, there's this new quantum update that is introducing a 3rd probability function, but it has to be compiled back down to the legacy 1.0 data types to get any practical use out of it at the moment.
Just want to offer a different perspective where the article author and Prime neglected all the time. Ironically, all the long live things the author worked on based on the article seems like not user-interaction heavy. That why they can write code in such a stable environment. In fact, if you are backend dev, you pretty much will have the same feeling. While web is aggressively changing over the last 30 years, the entire environment, the device, the mobile, different browsers, the user expectation. From static, read only articles to interactive elements. Tons of changes. That's why all the frameworks have to keep up. This is not web exclusive, it happens to all programs that heavily involves "UI" and "UX“. And guess what, web is the only one stays long live. All my mobile apps, desktop apps can quickly be obsolete if there is a minor change from the platform side. A old website from 15 years can still be open by a latest browser without issues. This is not trying to say web is good, we could do better, but it did perform pretty well til today.
Taking a 10 year break from JS and coming back to it is like coming back from the Thanos blip. The part who they were is still there but how they operate and their relationships are all different now.
It's basically my state lol, i've 'learned' a bit of html/php/javascript back around 2003, and now, i see what you actually need to do something is just horrible and depressing
I did a 2 year break from web development and let me tell you this. I don't want to be a web developer either. It was nice knowing handlebars, express, node, js etc.. but I - am - out .
Bro, you are the coolest technician ever. To write a senior capstone project on Rodney Mullen, that’s gnarly AF. I transitioned from skateboard to Web Developer at the graduating college. I know what you mean by pivoting
My first live professional system (for a department of the UK Civil Service) went live in 1982. I moved on from software development in 2005. Now getting back with Rust & htmx. Very happy to have jumped the era of huge JavaScript frameworks 🤣🤣🤣🎉🎉
When you write "rust &htmx" do you mean these are two tools that interest you or are you specifically saying youre using them together. I'm super ignorant of Rust, but isn't it a little much for building a MVC web app?
@@GabrielRivera-wx9sk yup I mean using Rust and htmx together. Rust overkill - maybe but the feature set does promise lowest hosting costs with maximum speed and reliability. Toolset is nice too.
The guy who wrote this article definitely made a lot of good points, but the sentence "I don't use many open-source packages beyond Swift math libraries, and most of the code I write is not something you would find anywhere" makes this guy a walking red flag on any team. Tbh it partially ruined the article for me. I don't want to work with someone who is so sure of themselves that they'll never use open source libs that aren't their own creation. This guy's condescension is off the charts.
as the guy did not really mention why - my own take on it is yet another dependency and for some things it makes sense to incorporate some of the features used part of the in-house framework rather than adding another dependency and all this entails
He kind of resembles that two faced guy who s lying and it takes a while of working with them to actually see that what they say dosen't align with what they do most of the times. So while everything here seems valid and some points are good, I doubt there s nothing else involved in this that he omits on purpose.
This is why I am so excited by HTMX. Pick a back end language and then use HTMX for the front end. Job done. Sure it won't be suitable in all cases but you could augment it with a sprinkling of front end javascript for edge cases if you really must. Hell, you could even rewind the clock and go back to PHP if you really wanted to, with the benefit of super cheap hosting like the old days. But I'm thinking Go with HTMX because of the performance, simplicity and standard library for templating and routing.
im currently learning go as a frontend dev mostly out of frustration with the field, it genuinely feels like frontend is just endless solving of invented problems and i want out of here asap. go with htmx looks super promising
Got me good on the "Sorry but I don't want to build a route handler" ... "actually im going to write that down and I'm going to build a route handler". I just stood up a "mono"-lambda in Go and built my own route handler (passed via API Gateway) so as not to have to bring in libraries like gorilla or gin. Small performance improvement, sure. Took a few days instead of an hour, sure. But great learning experience
I started professional web dev in 1998 at my first job and also left it in 2009. When I returned in 2013 I basically had to relearn everything from the ground up... it was a completely different world.
7:48 I actually quite like Borland back in the day. It's what I used for my AP CS101 back in high school. I thought their C/C++ compiler was really fast. It compiled the same source with full optimizations on faster than mingw gcc with -O0 opts disabled which I thought was impressive..
I use to have that mindset of everything should be webforms or csharp but i realized I need to add more experience to my workforce. React has become my heart of choice but i can easily adapt to csharp code to support my backend needs.
There's a lot of hubris in this article - but the lesson for all when doing something - pick the stack that has stood the test of time with the least moving parts, not the latest whizzbang. After all, they're just a bunch of if's and loops anyway, let's not kid ourselves.
@@S-we2gp The test of time I would say is 5+ years, AND a long-term track record like java or c# or python, and preferably a big player as a backer like Microsoft.
To that point about rewrites, I think that's something the industry really needs to get better at. You're entirely right. We saddle ourselves with years of legacy, and then someone comes along with the bright idea to not saddle ourselves with legacy, the rewrite gets mismanaged, somebody's head rolls, and then there's whole teams of devs that have to try to learn the legacy system again. Either that or the company goes under. Somehow the whole industry needs to figure out how to start shipping rewrites or how to not build balls of mud.
hello everyone, i need serious advice!, I am quite an Anticipatory person, I really like to be stable and changes can be quite frustrating. i started my web development journey 6 months ago and I have only learned (HTML,CSS and javascript with git and github), yes I am a slow learner too. I get very demotivated from the type of videos about "AI taking over" , so should I stay in this website development field or should I try something else. about me: I am 17 I started web dev just because I was bored
Try something else maybe adjacent to tech like project management or something. If you are a slow learner and like stability this likely isn’t the field for you.
Learning stuff for the first time is going to feel a lot slower than people who already know the core concepts of programming jumping on all the latest frameworks. 6 months is barely anything. The real question is : Do you enjoy it? What part about it do you enjoy? If you enjoy it, then you're not wasting your time. Regardless of what comes of AI. Just like art or carpentry or whatever else, it's fun to build. Do you enjoy the process of building the UI or the scripts or other parts?
I'm a slow learner as well, don't give up. I have also learnt only html, css and javascript. Don't listen to any negative thoughts. If AI kills web dev, then it will also kill data analysts, lawyers, general doctors ,teachers, etc. Just build small apps like calculator, clock, crud, shopping cart, tmdb api app, and then move on to react and make full clones
@@rayyanabdulwajid7681 The tech industry can be brutal. This isn’t a field that I would encourage anyone to follow who is self admittedly slow at learning things and also likes stability. These two personality traits are completely at odds with how professional software development works. There are many other professions that this person would be much more suited for so there’s no reason to force it.
When writing software, developer, customer/stakeholder should be aware or life cycle and required maintenance. Rewriting software should not became surprise if it is designed to be short lived. Of course developer can choose what technology to use make software match life cycle requirement.
The author has a completely different (maybe more sane) mindset than most webdevs. I don't think most web apps are developed with 5 years let alone 20 years of lifetime in mind.
Having a lot of fun as a webdev. Building your own stuff from scratch is just awesome. Also the vast ocean of options adds to the fun as there's always something to learn and improve upon. Constant learning is one thing I love about being a programmer... Sounds like someone who's miserable that a new patch destroyed his favourite game, in his eyes.
The most valuable lesson I learned from an old man was two words. "code rots." it was not a lesson to learn over night, but many years. Even when I was making more conscious decisions, the code inevitably became pure chaos. I came to the realization that you either build the app from scratch and move on, or inherit all the problems. and what prime said is true, don't try to rewrite ancient code. you will get fired.
As a newbie, is application longevity never considered when writing an application? From my rudimentary understanding of C, if I use C89 for my code base then as long as I have a C89 compiler for the specific OS, it will work the same across OSs right? And the code doesn't need to be changed regardless of how newer versions of C gets updated right? I'm a web developer so I know nothing about this so feel free to correct me on anything
Started doing Webdev mid 90th and while I think that it is so much harder for new Devs today, I was happy with every single step, that help me build good user experience instead of debugging asynchron data loading or DOM manipulation etc.
This is why a refuse to invest time in frameworks that over optimize what is typically a small portion of an overall application. Maybe that's because I don't want to be a front-end developer, you don't get much credit for making the button look slightly better.
15:30 its NOT progression at all, i started coding in the 80s too and most of the stuff today in (web)development seems to be just another different approach to solve something in another inefficient way. Thats why the stuff gets abandoned so quickly, if it was any good, IT WOULD LAST longer. Frameworks in coding is the same like shitcoins in crypto, as long as there is a huge demand from the hype-loving audience the more hype-BS will continue flooding the market. Btw, some of the things you mentioned in regards to programming languages is misleading, convenience-features are not an indicator for progression, if you could have implement it before as a helper function thats a sure sign it doesnt count as a progression for the coding-language itfself. I still get most of my web sh*t done without any preprocessors and some huge build-pipeline, text-editor is still enough and breaking a huge task into manageable fragments seems to be a skill that got totaly lost, we have quiet the opposit, even trivial tasks get blown out of proportion, everyrhing seems to be bloatware these days. Yin&Yang, for the pro's of moores law on one side, every two years there is a new generation of coders that introduce coding practices that result in inefficiency and redundancies and arent even aware of it that they are counteracting moores law with inefficient solutions.
Well, I'm only using an IDE for Python, for my webite in completely vanilla web technologies I just use xed (the Linux Mint standard text editor) and an FTP access to the server.
Rodney Mullen was so much more than just ‘slick’ on a board. He’s the reason skateboarding is what it is today. He created basically every well known/classic skateboard trick. He’s started making skateboards because he didn’t like the style of the original skateboards and created the modern skateboard(with the curve and pop). He’s beyond influential. Every single pro skateboarder basically came after Rodney and wanted to be like him in one way or another. He’s hands-down the most significant skateboarder in skateboarding history.
The argument shouldn't be more innovation vs. less, but rather about the fundamentals. The article mentioned that javascript was a questionable foundation for these technologies at best, and I think that is the real insight. A lot of the "innovation" is really just trying to patch over the flawed from the start technologies underpinning all of web-development.
Javascript has nothing to do it, it's simple demand supply situation. Other programming languages have their own frameworks too, but web killed those apps, because of how easy it is to develop in browser. Solving this is impossible because it's asking for standardization, which won't happen when it evolves so fast.
You still need solid javascript understanding to use it's frameworks. Framework won't listen to your clients problems, but you do. You can't be good at react and angular without having solid fundamentals of javascript
as a java developer this is exactly my impression from the whole web framework scene, mos of the other stuff he mentioend like assembly html, have evolvd not changed, but with web frameworks, it's really feels like there's no end, a simple "hello world" projects gets generated with 10 frameworks for linters, package managment, ci, unit testing, css frameworks, webpacks and a lot of other stuff, currently i'm using gwt , because it's accessible and easy, but doing actual modern web stuff it's feels impossible to, get into. not to mention menu and forms layouts, if i want to be able to quickly prototype stuff a backend using a webframework, it's an issue.
Someone I'm not surprised in the least you like Rodney Mullen. You guys would have the chillest sit down beer together. Man that guy could dance on the board
I remember interviewing with a start-up back in late 2019 and these clowns were excitedly telling me that they were re-writing their entire app they wrote in Scala in Go.....and not one of them knew anything about Go (They hired a "Go consulant"....basically some clown who knew just enough about Go to impress them). Bullet dodged.
I was a web developer from 1995 to 2017. I decided to "retire" because I couldn't take sitting in a cubicle anymore. Fast forward to today, I considered returning. Unfortunately web development isn't like riding a bike. Seven years out of the game might as well be 100. I'm basically starting from scratch. Even looking at job postings makes my head spin. It seems they want you to know every JavaScript framework there is. When I left, I was mainly a php developer with a good understanding of vanilla JS. It's daunting. It's like waking from a coma to find that no one speaks English anymore.
Hi Derek. This is sort of like my problem, though I'm not experienced like you. I'm a beginner developer trying to get my first job but pretty intimidated and overwhelmed by the amount of stuff you apparently need to know to get an entry-level job. It's very frustrating.
14:03 Never thougth about it but summarizing my career: 1st job -> Rewrite pre 03 c++ skype ancestor p2p code into 11 for a crypto. 2nd job -> No rewrites here :) 3rd job -> Rewrite mfc into Qt for a pcb software. 4th job -> Rewrite python into C# for crypto trading platform. How is that infrequent?
I propose a ultimate test.. build a desktop app and a web app that do the same.. lets see which one will colapse first for some lib update, security issue or simply a browser update... My sense says to me that desktop app will be a lot more reliable and have more longevity.
Bro it’s crazy to say “I stopped doing web development and started doing mobile” like you are in a huge auditorium, you said “I’m tired of this seat!” Stood up and sat down in the seat directly next to you
What else is there? I think webdev is popular because you can immediately see what you're building and show it to others. I've been stuck doing web dev for years now I'm getting burnt out and ready to explore other stuff, but I genuinely don't know what else is out there? I've a book on python, but I don't really know what to build? My BiL is into C and low-level languages and he's suggested I try to follow along building a game but I don't think I'm smart enough for stuff like. My mind goes blank when it comes to building anything not webdev related and I wonder if I'm even smart enough to break out of the bubble.
@@JacoBluezz AI dev, machine learning dev, mobile dev, embedded dev, gamedev, simulation software dev, compiler dev, backend dev, system administrator, database dev...... the list is goes on.....
Nobody in ITs glorious past seemingly gave two hoots about security, future issues, or anything past 1986. Running software using a 40 year old compilation on old or new hardware seems asking for trouble, even more so if it has any path to the wild. For cases in point, see the Y2K issue, see banks with super-magic perl scripts, and the massive issue with COBOL. The packages change, you don't have to update, but you need to accept the risk.
i can't lie after my last gig in web dev finished i said never again lol. got tired of writing js and php. Shifting into cybersecurity and do some game dev in my free time to tickle my brain
10:47 Prime clearly just had enough at this point. Security has always been paramount in software. It just taken so long to really gain perspective on that need. Source: Any software that has killed anyone written before 2002...
I’m learning c++ at the moment to try help recover from 10 years of JS development and give me some hope to continue working in technology. If that doesnt work im doing a trade lol. Ps NodeJS native modules look cool ;)
Bro not gonna lie to you. Neither me. I want to be a Machine Learning Engineer but it is so hard to switch I just found my self keep accepting web development jobs.
Machine learning jobs are very less in number, because they require researchers to and experienced software engineers to work at algorithm level , not just building some lame face recognition system
I started my first web page in 1999 and I'm still a developer... got made redundant last month, not due to performance but company financial position. Keep in mind, I specialised which was good for awhile until now. Have applied for at least 30 jobs, received 1 interview and didn't get it. Completely deflating - considering taking a break and deciding between moving back to full-stack, starting a business, or becoming a youtube influencer.
the problem with the software industry is techbro entitlement, as a 37 yr developer the behaviour is out of control, quit hiring ppl under 30 they are not mature enough to handle both the responsibility and accountability needed to better society, there is lack of accountability, its is so bad getting hired is an excercize in futility, hr mommy is everywhere and will devalue any man, because her 18 yr old hs graduate needs to make 180./yr and spend 100 of it on diablo forget the men trying to raise families and lift others out of poverty, moar $$ for children who devalue and outsource you. us men in our 40s have been left defeated, no offspring, expected to raise other mens children, and expected to engineer our fate.
I just realized I've started learning programming 8 years ago. When did SQL add javascript functions? I guess I should go back and look at some of these older programing languages I first learned. They changed a lot more than I realized.
It is true though that the learning curve is just coming back again and again today. Before it used to be that you learn core Java or core C or PHP, and you wouldn't have to learn a new API or read a ton more documentation every month - you do that now. Now new things come out all the time and have their own documentations and commands. So it's like a never-ending treadmill.
It has been interesting working as an embedded software engineer for the last decade and a half. I have had mad respect for the web people for having to constantly learn new shit just to do the basic stuff, but now that IoT has become far more common in embedded I have to suffer for your sins, lol. But I think this growth is truly a good thing, not a bad one. We would not have half the nice things if we were all writing cathedral-style long-living apps in C for everything. I may have to wrap my head around crazy web tech, but I also get stuff like Rust, so it kind of evens out.
Controversial take but... it all stems from wanting to solve a necessity using what you had, no matter how unfit the tools were, how crap the delivery vehicle was and how unsuitable the canvas you painted on was. And it can obviously only end with a ever growing hodge-podge frankenmonster epic kludge. With technical debt up to 11.
The last serious web development I did was back when jQuery was first rolling out and making life relatively better. I dabbled a little with WebRTC around 2012 but not for long. Taking a look at modern web dev today is just absolutely baffling how overly complex it has become for no good reason.
If you feel overwhelmed by the feeling to need to learn every new tech immediately, maybe your desire for likes and approval is bigger than your desire to make a good tech decision.
Good old memories, I also build SPA round that time on the same way. I also think we make it too complex, and are too depending on large frameworks. Keep it simple, and it also works nice and fast.
The real problem is that people are massively underestimating the cost of changing the tools and languages and frameworks, and massively overestimating the benefits. In fact most of the time people want to switch and do something new they literally don't even attempt to estimate the costs, or examine any disadvantages that come with it. When the changes do bring problems those same people who pushed for change simply deny the problems exist or blame the associated problems on other people. A lot of the times people people are just randomly deciding to change tools and then MAKING UP imaginary problems to justify their decisions.
The first website I worked on was a CGI script written in Borne shell. The second site I worked on was an Angular 2 SPA. I was so fucking lost. I still am, but I was then too.
I think that the main problem is over-engineering! Only a FAANG needs clean code, solid, a lot of frameworks, patterns, microservices, tools, tech tendencies and more...because they have a big product. If you are a small, mid or big company you don't need it, because it's cheaper building a new app every 4 year.
It's not about things not changing, it's about things not breaking. I have a 10 year old webapp that I have never had to change because of a version upgrade, because the libraries and programming language I used care about backwards compatibility. I'd rather solve the problems of tomorrow rather than the problems of yesterday.
lol 😂😂😂 they don’t know anything about programming they just hot online and blog nonsense… what I know is whatever any programming language you choose.. you must get yourself updated all the time…
Honestly I never liked web dev from the start. I got the hell out of it back in maybe 2003 and to be fair I never really got IN to it to begin with. I was instantly turned off by all the BS you had to do to play nice with every different browser out there and how finnicky it was to even get anything to be consistent.
Devs : Microsoft your APIs cause carpal tunnel Microsoft : Our APIs are bundled with Microsoft Ergonomic keyboard and mouse (paid separately ) Devs : Microsoft is such a great company
times got better, where is the point using stadardized technologies along with vanilla js, webcomp, webassembly, new browser features making lot frameworks redundant
Necessity drives innovation .... but here, you might not agree, technologies like C#, .net, JVM, JRE, Java, PHP .... provide that stability. I have heard a few number of times that company so and so uses this and that mainly because of politics not because that is the best available tech on the market. Long term stability is a policy and if a company is making a long term project it will need long term support a feature that might not be guaranteed by smaller companies that provide innovative solutions. We can all agree that companies like Microsoft, Apple and Oracle will probably be here in the next 10 years, unless something spectacular happens, but that can't be said for a large number of others.
I find the tired old node modules rant so boring. Most of the shit in there is build pipeline stuff but the people who criticize the number of packages in node_modules always seem to think that your literal client bundle is composed of all that.
4:21 I think the "Lindy effect" is really powerful in this regard. Paradoxically, it's often the oldest stuff that is worth learning, that will stick around for longer than whatever brand new shiny thing has appeared at the cutting edge. If it's an old book and people are still talking about it, it's probably worth reading. It's not likely that someone is going to come along and completely shatter our foundational understanding of philosophy and mathematics. Not impossible, just unlikely.
Indeed, investing in old stuff that is still around and relevant is the conservative thing to do, but we always need to keep an eye on the new stuff.
Interestingly, you don’t mention physics. Because you know it’s been shattered by Einstein and Bohr 😉
@@EdwinMartin The Lindy effect is just an observation that whatever has survived a longer time is more probable to survive in the future relative to recently produced things. That doesn’t mean new stuff won’t replace old stuff.
@@EdwinMartin Arguably, no. Classical physics is still the foundation of physics in engineering. Newtonian mechanics is in no practical sense obsolete because of relativity or quantum mechanics.
Absolutely. Proven, battle-hardened tech is such for a reason.
1:13 I actually did take a long ass break from webdev a little after jQuery got popular, and coming home to see paint everywhere is a totally accurate description of what it felt like to find out about shit like Node and React.
I still don't "get" Node. Yes, the react pattern was cool. Yes, it was neat how you could start a web site with a tiny amount of code. Neither one was a Javascript--specific feature and Node is (IMHO) a bad implementation of these patterns.
Which gives rise to the supposed "advantages" of using the same language on the client and server. "Advantages" that have never materialized in the data I've collected on productivity. Not even once.
@@thewiirocks All node is, is an extraction of google's Javascript engine outside of the browser, so you can run Javascript in other environments, server side, desktop application etc. Never made much sense to me as I was never a fan of Javascript. But it was never aimed at me, it was aimed at all those people that only knew Javascript.
React went off the rails shortly after it was introduced.
I think that transpilation is an insane concept. We are picking interpreted languages and treating them like compiled languages. I'm glad that Rust exists now to fix this madness.
@@rumplstiltztinkersteinRust is on a rushing change also. But it has a strong typed system and a strong compiler. So everything is always kept on the rail of compatibility. That's really amazing.
13:45 I worked on a project for 4.5 years and rewrote it twice in that time, about 3 months each. The DB greatly changed in the first rewrite and was sharded in the second. Huge changes, other than the Excel library (Ignite UI) we used. Write your code like you intend to rewrite it every few years. I've lived by that since I first heard it and it works so well with the evolution of frameworks and architectural decisions.
It's indicative of slow pace & small size of your team. Anyone who has worked in a fast paced environment will confirm, that amount of code often so is large, that it's practically impossible to rewrite it. Not mentioned the political capital you need to pull that off. Additionally, in companies with high focus on personal performance (which is all top companies) such refactoring is detrimental to your career and can get you fired.
I'm not sure what you mean write your code like you intend to rewrite it every few years?
@@S-we2gp Mostly that I don't expect anything I write to be around forever. Technology changes, org needs change, life moves on from what you wrote. Even staying in the same place for 3 years, I rewrote everything once to improve performance. Maybe it's different for people not in Full Stack roles...
Building the first 80% from scratch, thats the really enjoyable part. Over the years, I've carefully positioned myself in my company to get in on a new project, build it up to almost maturity and then move on to the next interesting project. Working too much on the last 20% will lead to burnout, or at least that's what it feels like to me.
I made the decision to just switch to Rust and C++ for this very reason.
Yes those languages are hard but there's a sense of completion.
Also I couldn't care less about React or Next.js (It feels taboo to say that out loud)
Out of curiosity, what are you working with atm? I am also eager to escape web and Rust and C++ look very interesting to me but struggling to pick a field because I don't really want to do games.
@@thesaintseiya Solidity is my strongest language, even though I worked on JS / TS for longer.
I feel you on the game dev path. That's another pro of low level languages. Career flexibility: Games, VR, Robotics, CLI tools you name it.
Solidity is awesome BUT it is specific to blockchain and the Ethereum chain. If anything happens to Ethereum, you are screwed.
Right now Arbitrum are working on a tool that will make it possible to write contracts in C++ and Rust... That should tell you something.
So you use webasm? Or are you just ignoring the web altogether? Most people don't wanna do that.
@@hyperthreaded I may look into it. But that's a HUGE maybe. Just bored of web. 🤷♂️
Also WASM is pretty niche as JS dominates the web. I'd rather use Rust and C++ to do things that JavaScript can't do.
@@dezly-macauleyDid you managed to find job in Rust or c++ after being JS dev? How hard it was? How long it took?
3:40 We also have a lot of applications that are stuck on older package versions. Every time time a vulnerability gets discovered, we need to upgrade all applications, as it should be. But often enough the fix was applied in some next major version. Support for other than the latest version is often very poor, forcing you to upgrade. And since it is a major version upgrade, the solution is often non trivial. To keep up we need to have one persion that does nothing but constantly upgrading the applications we're responsible for.
This is definitively a major downside of a micro-service/frontend architecture
Mullen is a legend. No exaggeration at all. It took decades for freestyle like what he helped pioneer really start to be accepted back into the mainstream of skating many years after the whole X games phenomena.
The best skill in web development: Focusing on your foundation first and ignoring the hype of the new tools. Frameworks are great but only after your foundation is strong.
I know how he feels. Working on my personal site is a lot more fun than working for a corporation. My current site is my second attempt at an SPA, and the first that doesn't break the Back and Forward buttons. I write all of my code from scratch with no frameworks, borrowing snippets from Stack Overflow as needed.
9:50 You know what Hasn't changed that much... Binary. The language literally has only two functions. 1 = true/on and 0 = false/off. I mean, there's this new quantum update that is introducing a 3rd probability function, but it has to be compiled back down to the legacy 1.0 data types to get any practical use out of it at the moment.
Next time I fck up, I am just gonna say "I have a mind of an artist"
Just want to offer a different perspective where the article author and Prime neglected all the time.
Ironically, all the long live things the author worked on based on the article seems like not user-interaction heavy. That why they can write code in such a stable environment. In fact, if you are backend dev, you pretty much will have the same feeling. While web is aggressively changing over the last 30 years, the entire environment, the device, the mobile, different browsers, the user expectation. From static, read only articles to interactive elements. Tons of changes. That's why all the frameworks have to keep up.
This is not web exclusive, it happens to all programs that heavily involves "UI" and "UX“. And guess what, web is the only one stays long live. All my mobile apps, desktop apps can quickly be obsolete if there is a minor change from the platform side. A old website from 15 years can still be open by a latest browser without issues.
This is not trying to say web is good, we could do better, but it did perform pretty well til today.
Taking a 10 year break from JS and coming back to it is like coming back from the Thanos blip. The part who they were is still there but how they operate and their relationships are all different now.
Its like in the shawshank redemption, brooks being sent back into society after 50 years in prison.
It's basically my state lol, i've 'learned' a bit of html/php/javascript back around 2003, and now, i see what you actually need to do something is just horrible and depressing
I did a 2 year break from web development and let me tell you this. I don't want to be a web developer either.
It was nice knowing handlebars, express, node, js etc.. but I - am - out .
enjoy being poor and having a soul
What are you doing now?
OnlyFans creator
Bro, you are the coolest technician ever. To write a senior capstone project on Rodney Mullen, that’s gnarly AF. I transitioned from skateboard to Web Developer at the graduating college. I know what you mean by pivoting
My first live professional system (for a department of the UK Civil Service) went live in 1982.
I moved on from software development in 2005. Now getting back with Rust & htmx.
Very happy to have jumped the era of huge JavaScript frameworks 🤣🤣🤣🎉🎉
When you write "rust &htmx" do you mean these are two tools that interest you or are you specifically saying youre using them together. I'm super ignorant of Rust, but isn't it a little much for building a MVC web app?
@@GabrielRivera-wx9sk yup I mean using Rust and htmx together. Rust overkill - maybe but the feature set does promise lowest hosting costs with maximum speed and reliability. Toolset is nice too.
The guy who wrote this article definitely made a lot of good points, but the sentence "I don't use many open-source packages beyond Swift math libraries, and most of the code I write is not something you would find anywhere" makes this guy a walking red flag on any team. Tbh it partially ruined the article for me. I don't want to work with someone who is so sure of themselves that they'll never use open source libs that aren't their own creation. This guy's condescension is off the charts.
as the guy did not really mention why - my own take on it is yet another dependency and for some things it makes sense to incorporate some of the features used part of the in-house framework rather than adding another dependency and all this entails
He kind of resembles that two faced guy who s lying and it takes a while of working with them to actually see that what they say dosen't align with what they do most of the times. So while everything here seems valid and some points are good, I doubt there s nothing else involved in this that he omits on purpose.
Whenever I complain about JS I remember that Java was almost chosen as the standard web language.
Be verbose as it is, at least it's typed.
@@3_smh_3and has a decent standard library
Using Kotlin instead of JS would be verry nice. Java is not the only language that compiles to java byte code.
Java > JavaScript
JavaScript is terrible. Let, const, var, globals, undefined, null, use strict, a million module types... it's a mess
This is why I am so excited by HTMX. Pick a back end language and then use HTMX for the front end. Job done. Sure it won't be suitable in all cases but you could augment it with a sprinkling of front end javascript for edge cases if you really must. Hell, you could even rewind the clock and go back to PHP if you really wanted to, with the benefit of super cheap hosting like the old days. But I'm thinking Go with HTMX because of the performance, simplicity and standard library for templating and routing.
im currently learning go as a frontend dev mostly out of frustration with the field, it genuinely feels like frontend is just endless solving of invented problems and i want out of here asap. go with htmx looks super promising
I doubt if it can do proper dom manipulation like javascript
@@rayyanabdulwajid7681 Which is exactly why I said it won't be suitable in all cases.
Rodney Mullen single handedly invented skateboarding and the magic ollie
He's been my favourite since I discovered him (and skateboarding) as a little kid via tony hawk's pro skater 4 on ps1
I mean, dude was laid up with a broken foot when he figured out the impossible
I think he was a plant by the grip tape industry. Darkslides get outta here
@@cdb8987 🤣💯
@@cdb8987 watching his video on THPS2 back when I was a kid blew my damn mind
Got me good on the "Sorry but I don't want to build a route handler" ... "actually im going to write that down and I'm going to build a route handler". I just stood up a "mono"-lambda in Go and built my own route handler (passed via API Gateway) so as not to have to bring in libraries like gorilla or gin. Small performance improvement, sure. Took a few days instead of an hour, sure. But great learning experience
Makes sense! If you don't need the web server listener and ports, etc, then no need to bring in an entire package like gorilla-mux
Exactly! @@brandonwlee2013
Thank you for the awesome content. I am currently going through your course on frontend master (y)
I started professional web dev in 1998 at my first job and also left it in 2009. When I returned in 2013 I basically had to relearn everything from the ground up... it was a completely different world.
7:48 I actually quite like Borland back in the day. It's what I used for my AP CS101 back in high school. I thought their C/C++ compiler was really fast. It compiled the same source with full optimizations on faster than mingw gcc with -O0 opts disabled which I thought was impressive..
Custom frameworks are made for job security, those that no one understands except the guy who wrote it.
Ollie Gelfand - he now runs an import repair shop. I wrote articles for him and helped him with his website. 👍
I use to have that mindset of everything should be webforms or csharp but i realized I need to add more experience to my workforce. React has become my heart of choice but i can easily adapt to csharp code to support my backend needs.
There's a lot of hubris in this article - but the lesson for all when doing something - pick the stack that has stood the test of time with the least moving parts, not the latest whizzbang.
After all, they're just a bunch of if's and loops anyway, let's not kid ourselves.
but in web whats the latest whizzbang 1 year old? whats the test of time 5 years, 8 years? The web has such a weird time scale
@@S-we2gp The test of time I would say is 5+ years, AND a long-term track record like java or c# or python, and preferably a big player as a backer like Microsoft.
To that point about rewrites, I think that's something the industry really needs to get better at. You're entirely right. We saddle ourselves with years of legacy, and then someone comes along with the bright idea to not saddle ourselves with legacy, the rewrite gets mismanaged, somebody's head rolls, and then there's whole teams of devs that have to try to learn the legacy system again. Either that or the company goes under. Somehow the whole industry needs to figure out how to start shipping rewrites or how to not build balls of mud.
Can't believe there was no mention of Rodney vs Daewon, peak picnic table usage!!
hello everyone,
i need serious advice!, I am quite an Anticipatory person, I really like to be stable and changes can be quite frustrating. i started my web development journey 6 months ago and I have only learned (HTML,CSS and javascript with git and github), yes I am a slow learner too. I get very demotivated from the type of videos about "AI taking over" , so should I stay in this website development field or should I try something else.
about me:
I am 17
I started web dev just because I was bored
Try something else maybe adjacent to tech like project management or something. If you are a slow learner and like stability this likely isn’t the field for you.
@@relaxingsounds5469 thanks for your response, means alot to me I'll take it into consideration moving forward
Learning stuff for the first time is going to feel a lot slower than people who already know the core concepts of programming jumping on all the latest frameworks. 6 months is barely anything.
The real question is : Do you enjoy it? What part about it do you enjoy?
If you enjoy it, then you're not wasting your time. Regardless of what comes of AI. Just like art or carpentry or whatever else, it's fun to build.
Do you enjoy the process of building the UI or the scripts or other parts?
I'm a slow learner as well, don't give up. I have also learnt only html, css and javascript. Don't listen to any negative thoughts. If AI kills web dev, then it will also kill data analysts, lawyers, general doctors ,teachers, etc. Just build small apps like calculator, clock, crud, shopping cart, tmdb api app, and then move on to react and make full clones
@@rayyanabdulwajid7681 The tech industry can be brutal. This isn’t a field that I would encourage anyone to follow who is self admittedly slow at learning things and also likes stability. These two personality traits are completely at odds with how professional software development works. There are many other professions that this person would be much more suited for so there’s no reason to force it.
When writing software, developer, customer/stakeholder should be aware or life cycle and required maintenance. Rewriting software should not became surprise if it is designed to be short lived.
Of course developer can choose what technology to use make software match life cycle requirement.
The author has a completely different (maybe more sane) mindset than most webdevs. I don't think most web apps are developed with 5 years let alone 20 years of lifetime in mind.
Laravel and htmx with go microservice sidecars is oretty great for many cases.
The problem is if there is a real company behind it it stays so long only some logo and UI updates
Well they can’t be. Things change too much and budgets aren’t there to build robust software like what would be seen running mission critical systems
Having a lot of fun as a webdev. Building your own stuff from scratch is just awesome. Also the vast ocean of options adds to the fun as there's always something to learn and improve upon. Constant learning is one thing I love about being a programmer...
Sounds like someone who's miserable that a new patch destroyed his favourite game, in his eyes.
The most valuable lesson I learned from an old man was two words. "code rots." it was not a lesson to learn over night, but many years. Even when I was making more conscious decisions, the code inevitably became pure chaos. I came to the realization that you either build the app from scratch and move on, or inherit all the problems. and what prime said is true, don't try to rewrite ancient code. you will get fired.
As a newbie, is application longevity never considered when writing an application? From my rudimentary understanding of C, if I use C89 for my code base then as long as I have a C89 compiler for the specific OS, it will work the same across OSs right? And the code doesn't need to be changed regardless of how newer versions of C gets updated right? I'm a web developer so I know nothing about this so feel free to correct me on anything
i loved "don't package shame" LMAO
Started doing Webdev mid 90th and while I think that it is so much harder for new Devs today, I was happy with every single step, that help me build good user experience instead of debugging asynchron data loading or DOM manipulation etc.
This is why a refuse to invest time in frameworks that over optimize what is typically a small portion of an overall application. Maybe that's because I don't want to be a front-end developer, you don't get much credit for making the button look slightly better.
15:30 its NOT progression at all, i started coding in the 80s too and most of the stuff today in (web)development seems to be just another different approach to solve something in another inefficient way. Thats why the stuff gets abandoned so quickly, if it was any good, IT WOULD LAST longer.
Frameworks in coding is the same like shitcoins in crypto, as long as there is a huge demand from the hype-loving audience the more hype-BS will continue flooding the market.
Btw, some of the things you mentioned in regards to programming languages is misleading, convenience-features are not an indicator for progression, if you could have implement it before as a helper function thats a sure sign it doesnt count as a progression for the coding-language itfself.
I still get most of my web sh*t done without any preprocessors and some huge build-pipeline, text-editor is still enough and breaking a huge task into manageable fragments seems to be a skill that got totaly lost, we have quiet the opposit, even trivial tasks get blown out of proportion, everyrhing seems to be bloatware these days.
Yin&Yang, for the pro's of moores law on one side, every two years there is a new generation of coders that introduce coding practices that result in inefficiency and redundancies and arent even aware of it that they are counteracting moores law with inefficient solutions.
Well, I'm only using an IDE for Python, for my webite in completely vanilla web technologies I just use xed (the Linux Mint standard text editor) and an FTP access to the server.
Rodney Mullen was so much more than just ‘slick’ on a board. He’s the reason skateboarding is what it is today. He created basically every well known/classic skateboard trick. He’s started making skateboards because he didn’t like the style of the original skateboards and created the modern skateboard(with the curve and pop). He’s beyond influential. Every single pro skateboarder basically came after Rodney and wanted to be like him in one way or another. He’s hands-down the most significant skateboarder in skateboarding history.
The argument shouldn't be more innovation vs. less, but rather about the fundamentals. The article mentioned that javascript was a questionable foundation for these technologies at best, and I think that is the real insight. A lot of the "innovation" is really just trying to patch over the flawed from the start technologies underpinning all of web-development.
Javascript has nothing to do it, it's simple demand supply situation. Other programming languages have their own frameworks too, but web killed those apps, because of how easy it is to develop in browser. Solving this is impossible because it's asking for standardization, which won't happen when it evolves so fast.
You still need solid javascript understanding to use it's frameworks. Framework won't listen to your clients problems, but you do. You can't be good at react and angular without having solid fundamentals of javascript
as a java developer this is exactly my impression from the whole web framework scene, mos of the other stuff he mentioend like assembly html, have evolvd not changed, but with web frameworks, it's really feels like there's no end, a simple "hello world" projects gets generated with 10 frameworks for linters, package managment, ci, unit testing, css frameworks, webpacks and a lot of other stuff, currently i'm using gwt , because it's accessible and easy, but doing actual modern web stuff it's feels impossible to, get into. not to mention menu and forms layouts, if i want to be able to quickly prototype stuff a backend using a webframework, it's an issue.
Menu and form layouts are easy, and I'm still in the process of learning web development
jQuery was 2006, i made SPA in 2006 with jQuery and a homemade template system, that was the days, before most other devs too.
Great job, boy!
11:49
What does he mean by named template strings?
Programming on an IBM As/400 in 2023. Grab some code written in 1964, compile and run. No problem.
Job security, I hope.
As a sk8erboi in hogh school from 00 until i left for the army in 05 and wasnt allowed to skate anymore rodney mullen is the goat fosho
Someone I'm not surprised in the least you like Rodney Mullen. You guys would have the chillest sit down beer together. Man that guy could dance on the board
10:03 PHP *does* change.... a *lot* ! Speacially in tiny undocumented details that you will only notice months after upgrading
I remember interviewing with a start-up back in late 2019 and these clowns were excitedly telling me that they were re-writing their entire app they wrote in Scala in Go.....and not one of them knew anything about Go (They hired a "Go consulant"....basically some clown who knew just enough about Go to impress them). Bullet dodged.
I was a web developer from 1995 to 2017. I decided to "retire" because I couldn't take sitting in a cubicle anymore.
Fast forward to today, I considered returning. Unfortunately web development isn't like riding a bike. Seven years out of the game might as well be 100. I'm basically starting from scratch. Even looking at job postings makes my head spin. It seems they want you to know every JavaScript framework there is. When I left, I was mainly a php developer with a good understanding of vanilla JS. It's daunting. It's like waking from a coma to find that no one speaks English anymore.
Hi Derek. This is sort of like my problem, though I'm not experienced like you. I'm a beginner developer trying to get my first job but pretty intimidated and overwhelmed by the amount of stuff you apparently need to know to get an entry-level job. It's very frustrating.
14:03 Never thougth about it but summarizing my career:
1st job -> Rewrite pre 03 c++ skype ancestor p2p code into 11 for a crypto.
2nd job -> No rewrites here :)
3rd job -> Rewrite mfc into Qt for a pcb software.
4th job -> Rewrite python into C# for crypto trading platform.
How is that infrequent?
Still alive and typing?
I propose a ultimate test.. build a desktop app and a web app that do the same.. lets see which one will colapse first for some lib update, security issue or simply a browser update... My sense says to me that desktop app will be a lot more reliable and have more longevity.
I did Tibco General Interface in 2007, thankfully moved to Google Web Toolkit... Does any of these two count as framework?
Bro it’s crazy to say “I stopped doing web development and started doing mobile” like you are in a huge auditorium, you said “I’m tired of this seat!” Stood up and sat down in the seat directly next to you
Prime is always receiving beverages... jealous
its sad that most new devs think web dev is all there is out here. You don't have to be miserable web dev if you don't want to
What else is there? I think webdev is popular because you can immediately see what you're building and show it to others. I've been stuck doing web dev for years now I'm getting burnt out and ready to explore other stuff, but I genuinely don't know what else is out there? I've a book on python, but I don't really know what to build? My BiL is into C and low-level languages and he's suggested I try to follow along building a game but I don't think I'm smart enough for stuff like. My mind goes blank when it comes to building anything not webdev related and I wonder if I'm even smart enough to break out of the bubble.
@@JacoBluezz i want to eventually pivot to sth like gpu drivers or embedded, idk how feasible it is but it sounds way better than doing web
@@JacoBluezz AI dev, machine learning dev, mobile dev, embedded dev, gamedev, simulation software dev, compiler dev, backend dev, system administrator, database dev...... the list is goes on.....
Nobody in ITs glorious past seemingly gave two hoots about security, future issues, or anything past 1986. Running software using a 40 year old compilation on old or new hardware seems asking for trouble, even more so if it has any path to the wild. For cases in point, see the Y2K issue, see banks with super-magic perl scripts, and the massive issue with COBOL.
The packages change, you don't have to update, but you need to accept the risk.
Went from react to ue5.
👍
Working at a game company ?
yessss
Was react that difficult for you?😂 Also don't you know that the opportunities are greater in web dev compared to game development.
Oh, I admire you did that!
i can't lie after my last gig in web dev finished i said never again lol. got tired of writing js and php. Shifting into cybersecurity and do some game dev in my free time to tickle my brain
`?.` isn't the Elvis operator. `?:` is the Elvis operator. `?.` is the optional chaining operator.
10:47 Prime clearly just had enough at this point.
Security has always been paramount in software. It just taken so long to really gain perspective on that need.
Source: Any software that has killed anyone written before 2002...
Hey Prime, what's the thing you use in the browser where you can make mindmaps etc, saw it in a video but can't find it now.
Excalidraw probably
yep thanks :)@@livingonvideo
Desktop apps are my passion.
I fucking hate web!
I like networking and low level shit but there's zero money in it
but boy would I be happier to make c/++ apps
I’m learning c++ at the moment to try help recover from 10 years of JS development and give me some hope to continue working in technology. If that doesnt work im doing a trade lol.
Ps NodeJS native modules look cool ;)
Rodney Mullen is an incredibly inspiring public speaker now, too. You should look up his ted talk, "pop an ollie and innovate"
Bro not gonna lie to you. Neither me. I want to be a Machine Learning Engineer but it is so hard to switch I just found my self keep accepting web development jobs.
Machine learning jobs are very less in number, because they require researchers to and experienced software engineers to work at algorithm level , not just building some lame face recognition system
Get some ML books and study and train hard then
About React Andys: it does not take much to switch to Vue, Svelte, Solid or whatever from React. It's all the same, it's just written differently.
I started my first web page in 1999 and I'm still a developer... got made redundant last month, not due to performance but company financial position. Keep in mind, I specialised which was good for awhile until now. Have applied for at least 30 jobs, received 1 interview and didn't get it. Completely deflating - considering taking a break and deciding between moving back to full-stack, starting a business, or becoming a youtube influencer.
the problem with the software industry is techbro entitlement, as a 37 yr developer the behaviour is out of control, quit hiring ppl under 30 they are not mature enough to handle both the responsibility and accountability needed to better society, there is lack of accountability, its is so bad getting hired is an excercize in futility, hr mommy is everywhere and will devalue any man, because her 18 yr old hs graduate needs to make 180./yr and spend 100 of it on diablo forget the men trying to raise families and lift others out of poverty, moar $$ for children who devalue and outsource you. us men in our 40s have been left defeated, no offspring, expected to raise other mens children, and expected to engineer our fate.
@@invalidaccount2315 ... Every adult should be expected to raise every child and engineer fate, that's maturity.
I just realized I've started learning programming 8 years ago.
When did SQL add javascript functions? I guess I should go back and look at some of these older programing languages I first learned. They changed a lot more than I realized.
MySQL, not SQL
@@defeqel6537 potato tomato.
and a database stores data, not data.
It is true though that the learning curve is just coming back again and again today.
Before it used to be that you learn core Java or core C or PHP, and you wouldn't have to learn a new API or read a ton more documentation every month - you do that now. Now new things come out all the time and have their own documentations and commands. So it's like a never-ending treadmill.
This article comes off as someone trying really hard to play the cool kid.
i can sumpathise, I also transitioned from a webdeveloper to doing low level software development
It has been interesting working as an embedded software engineer for the last decade and a half. I have had mad respect for the web people for having to constantly learn new shit just to do the basic stuff, but now that IoT has become far more common in embedded I have to suffer for your sins, lol.
But I think this growth is truly a good thing, not a bad one. We would not have half the nice things if we were all writing cathedral-style long-living apps in C for everything.
I may have to wrap my head around crazy web tech, but I also get stuff like Rust, so it kind of evens out.
Controversial take but... it all stems from wanting to solve a necessity using what you had, no matter how unfit the tools were, how crap the delivery vehicle was and how unsuitable the canvas you painted on was. And it can obviously only end with a ever growing hodge-podge frankenmonster epic kludge. With technical debt up to 11.
The last serious web development I did was back when jQuery was first rolling out and making life relatively better. I dabbled a little with WebRTC around 2012 but not for long.
Taking a look at modern web dev today is just absolutely baffling how overly complex it has become for no good reason.
If you feel overwhelmed by the feeling to need to learn every new tech immediately, maybe your desire for likes and approval is bigger than your desire to make a good tech decision.
Good old memories, I also build SPA round that time on the same way.
I also think we make it too complex, and are too depending on large frameworks.
Keep it simple, and it also works nice and fast.
Lisp has changed from the 1950s to this day, and keeps changing
I wish I would have kept my Rodney Mullen board from the 80's...
The real problem is that people are massively underestimating the cost of changing the tools and languages and frameworks, and massively overestimating the benefits. In fact most of the time people want to switch and do something new they literally don't even attempt to estimate the costs, or examine any disadvantages that come with it. When the changes do bring problems those same people who pushed for change simply deny the problems exist or blame the associated problems on other people. A lot of the times people people are just randomly deciding to change tools and then MAKING UP imaginary problems to justify their decisions.
The first website I worked on was a CGI script written in Borne shell. The second site I worked on was an Angular 2 SPA. I was so fucking lost. I still am, but I was then too.
I think that the main problem is over-engineering! Only a FAANG needs clean code, solid, a lot of frameworks, patterns, microservices, tools, tech tendencies and more...because they have a big product. If you are a small, mid or big company you don't need it, because it's cheaper building a new app every 4 year.
It's not about things not changing, it's about things not breaking. I have a 10 year old webapp that I have never had to change because of a version upgrade, because the libraries and programming language I used care about backwards compatibility.
I'd rather solve the problems of tomorrow rather than the problems of yesterday.
Loved the Rodney Mullen namedrop. He's definitely the man :D
forgot the " | Prime Reacts"
I'm also tired of the web. I was looking at videos from "tech influencers" that were recommending the "best stack" for 2024. Please, stop.
lol 😂😂😂 they don’t know anything about programming they just hot online and blog nonsense… what I know is whatever any programming language you choose.. you must get yourself updated all the time…
have to love the "back in my day"
Honestly I never liked web dev from the start. I got the hell out of it back in maybe 2003 and to be fair I never really got IN to it to begin with. I was instantly turned off by all the BS you had to do to play nice with every different browser out there and how finnicky it was to even get anything to be consistent.
6:18 yeah, this is really what programming means
Devs : Microsoft your APIs cause carpal tunnel
Microsoft : Our APIs are bundled with Microsoft Ergonomic keyboard and mouse (paid separately )
Devs : Microsoft is such a great company
React is to the web what MFC was to windows desktop
times got better, where is the point using stadardized technologies along with vanilla js, webcomp, webassembly, new browser features making lot frameworks redundant
The HR meeting was hillarious 😂
Necessity drives innovation .... but here, you might not agree, technologies like C#, .net, JVM, JRE, Java, PHP .... provide that stability. I have heard a few number of times that company so and so uses this and that mainly because of politics not because that is the best available tech on the market. Long term stability is a policy and if a company is making a long term project it will need long term support a feature that might not be guaranteed by smaller companies that provide innovative solutions. We can all agree that companies like Microsoft, Apple and Oracle will probably be here in the next 10 years, unless something spectacular happens, but that can't be said for a large number of others.
I find the tired old node modules rant so boring. Most of the shit in there is build pipeline stuff but the people who criticize the number of packages in node_modules always seem to think that your literal client bundle is composed of all that.
immediately smashed that like button when you said Rodney Mullen, even before watching the rest of the vid
16:40 When Gen X finally does start to speak, you know it'll be some Be Not Afraid shit.