3 ms·

Just wanted to say that I appreciate this reply. I'm probably going to fall asleep soon, but FWIW we're both pretty much on the same page. Regarding "embrace t

Just wanted to say that I appreciate this reply. I'm probably going to fall asleep soon, but FWIW we're both pretty much on the same page. Regarding "embrace the suck," the best I can say while being brief is... Ah man, this is such an interesting topic, and it's worthy of a blog post in its own right. You have to trust me when I say "there's something here; it's worth tugging at this thread to find out whether it might be true." It is counterintuitive. But the fact that it's counterintuitive is a hint that it might be worth checking whether it might be true. Relativity was counterintuitive too. If you embrace the suck, you can pull the simplicity out of the chaos. But only after you've mastered the chaos. That sounds like some combination of naive or impractical. Who could possibly master chaos? You can. And that's all I'm saying. Step one is to force yourself to choose to try.

mercer 9y ago
I'm not sure I completely understand what you're getting at, but FWIW despite the many problems I'm incredibly excited about the front-end JS ecosystem. I do think it's a great situation where one single programming language allows one to not only create complete applications that run on every computer under the sun (without installing Java), but allow you to fucking debug it in a pleasant way on pretty much each of those computers. But all that said, I wish I hadn't started out with JS and blindly followed the 'chaos'. I've wasted so much time using React and Redux when simply React, or in some cases even good old Backbone of jQuery would have sufficed. I wish I'd learned good programming practices from more overtly functional or perhaps more overtly OO languages, before learning JS. I wish I hadn't spent many hours dealing with Grunt, no, Gulp, no Webpack, no Webpack 2, only to become a 'master' at something that is entirely irrelevant outside the JS ecosystem. Sure, I learned the intricacies of the 'chaos' and I find myself going more and more for simpler solutions within the ecosystem. I'm comfortable using Baobab.js or MobX instead of Redux. I'm comfortable using only lodash instead of a whole bunch of modules that do part of what lodash does. But I don't feel any of that made me better as a programming. In fact, as the JS ecosystem is improving, I find that a lot of my arcane knowledge of how to, say, enable hot code reloading in Webpack 1 is entirely useless now that Webpack 2 sort of does it out of the box. More and more I'm inclined to let the JS ecosystem do its crazy shit and wait for something semi-standard to shake out before I even bother. Because everything I learn about the specifics of this chaos will not be worth knowing in <x> months.
- sillysaurus3 9y ago
  Aha. I see the crux. You're right, in a particular way that I hadn't thought of: Most people don't have a background in programming. They weren't doing it when they were teens, and many people probably started writing code within the last year or so. In that context, I completely agree. It's absolutely true that at the end of all that chaos, you won't end up a better programmer. I was speaking as someone who was in the inverse situation: I'd amassed a lot of theoretical knowledge, and spent a lot of time chasing the phantom of "being a good programmer." I researched memory models, read whitepapers, explored how Google implemented bigtable, went through rtm's 6.824 Distributed Systems course just for fun (it's freely available online)... You know, a bunch of hardcore stuff that it seems like "real programmers ought to know." Yeah... Turns out that doesn't always translate into dollars. It came really close to being a bad situation (explained here https://news.ycombinator.com/item?id=15642428 https://news.ycombinator.com/item?id=15642428). Let's put it this way. If I had to do it all over again, it's very possible I would swap skillsets with you. Because you know all of the stuff you rattled off. You are now prepared to avoid it. So yes, it's easy to feel like you've wasted your time. But when it comes to employability, you are far better positioned from a career standpoint. Stuff like this may feel "out of bounds" of traditional conversation. Normally we're supposed to make up arguments that support our points purely through reason alone, right? Like it would be much more persuasive for me to say "Well, this matters because X" where X is something related to programming. Talking about career stuff and employability may seem like punting. It might also seem like it's something you don't really need to worry too much about. But when you relax and let yourself stop learning, and stop going through the motions of all that stuff you hated, it's merely 3-5 years before you end up in a similar position. Maybe. Or maybe you'll get lucky.
  - mercer 9y ago
    I completely agree, FWIW. I make a very decent living as a JS front-end dev using framework-du-jour. Thankfully that very often is React/Redux/React-Redux/React-Router, and a smattering of other React/Redux type stuff. But that doesn't make it less shit. I'm actively re-schooling myself as a more backend-ish, or perhaps full-stack-ish developer that allows me to use more of Elixir, Clojure, or even Ruby/Python, and less of JavaScript. I'd say there are basically three perspectives on this: 1. development ease/quality, where JavaScript is not the worst, but far from the best. 2. ecosystem health and best practices, where I find that JS is not doing a great job. A lot of churn, a lot of reinventing the wheel, and a lot of stockholm syndrome defending the mess. 3. money, where being a good front-ender is worth most of the effort and frustration. I'd say 3 is the most contentious. Most people seem to agree on 1 that JS and its ecosystem are not a paragon of good development. It takes work to even just get a good initial setup, which is not a problem in many other ecosystems. Transpiling, to name one example, is just not necessary. I think a sizable portion of multi-lingual devs would agree on 2 as well. But 3 is difficult to quantify. Is it a good thing that I can me a good chunk of money building React/Redux apps when I could've built a saner, more stable version in Django/Rails/Elixir in a fraction of the time if the only downside is that it requires a full page reload? I'm not so sure anymore. But it's worth a discussion. Point 1 and 2 strike me as obvious enough that bringing them up ad nauseum on HN is just pointless and frustrating.

mercer 9y ago

I'm not sure I completely understand what you're getting at, but FWIW despite the many problems I'm incredibly excited about the front-end JS ecosystem. I do think it's a great situation where one single programming language allows one to not only create complete applications that run on every computer under the sun (without installing Java), but allow you to fucking debug it in a pleasant way on pretty much each of those computers. But all that said, I wish I hadn't started out with JS and blindly followed the 'chaos'. I've wasted so much time using React and Redux when simply React, or in some cases even good old Backbone of jQuery would have sufficed. I wish I'd learned good programming practices from more overtly functional or perhaps more overtly OO languages, before learning JS. I wish I hadn't spent many hours dealing with Grunt, no, Gulp, no Webpack, no Webpack 2, only to become a 'master' at something that is entirely irrelevant outside the JS ecosystem. Sure, I learned the intricacies of the 'chaos' and I find myself going more and more for simpler solutions within the ecosystem. I'm comfortable using Baobab.js or MobX instead of Redux. I'm comfortable using only lodash instead of a whole bunch of modules that do part of what lodash does. But I don't feel any of that made me better as a programming. In fact, as the JS ecosystem is improving, I find that a lot of my arcane knowledge of how to, say, enable hot code reloading in Webpack 1 is entirely useless now that Webpack 2 sort of does it out of the box. More and more I'm inclined to let the JS ecosystem do its crazy shit and wait for something semi-standard to shake out before I even bother. Because everything I learn about the specifics of this chaos will not be worth knowing in <x> months.

sillysaurus3 9y ago
Aha. I see the crux. You're right, in a particular way that I hadn't thought of: Most people don't have a background in programming. They weren't doing it when they were teens, and many people probably started writing code within the last year or so. In that context, I completely agree. It's absolutely true that at the end of all that chaos, you won't end up a better programmer. I was speaking as someone who was in the inverse situation: I'd amassed a lot of theoretical knowledge, and spent a lot of time chasing the phantom of "being a good programmer." I researched memory models, read whitepapers, explored how Google implemented bigtable, went through rtm's 6.824 Distributed Systems course just for fun (it's freely available online)... You know, a bunch of hardcore stuff that it seems like "real programmers ought to know." Yeah... Turns out that doesn't always translate into dollars. It came really close to being a bad situation (explained here https://news.ycombinator.com/item?id=15642428 https://news.ycombinator.com/item?id=15642428). Let's put it this way. If I had to do it all over again, it's very possible I would swap skillsets with you. Because you know all of the stuff you rattled off. You are now prepared to avoid it. So yes, it's easy to feel like you've wasted your time. But when it comes to employability, you are far better positioned from a career standpoint. Stuff like this may feel "out of bounds" of traditional conversation. Normally we're supposed to make up arguments that support our points purely through reason alone, right? Like it would be much more persuasive for me to say "Well, this matters because X" where X is something related to programming. Talking about career stuff and employability may seem like punting. It might also seem like it's something you don't really need to worry too much about. But when you relax and let yourself stop learning, and stop going through the motions of all that stuff you hated, it's merely 3-5 years before you end up in a similar position. Maybe. Or maybe you'll get lucky.
- mercer 9y ago
  I completely agree, FWIW. I make a very decent living as a JS front-end dev using framework-du-jour. Thankfully that very often is React/Redux/React-Redux/React-Router, and a smattering of other React/Redux type stuff. But that doesn't make it less shit. I'm actively re-schooling myself as a more backend-ish, or perhaps full-stack-ish developer that allows me to use more of Elixir, Clojure, or even Ruby/Python, and less of JavaScript. I'd say there are basically three perspectives on this: 1. development ease/quality, where JavaScript is not the worst, but far from the best. 2. ecosystem health and best practices, where I find that JS is not doing a great job. A lot of churn, a lot of reinventing the wheel, and a lot of stockholm syndrome defending the mess. 3. money, where being a good front-ender is worth most of the effort and frustration. I'd say 3 is the most contentious. Most people seem to agree on 1 that JS and its ecosystem are not a paragon of good development. It takes work to even just get a good initial setup, which is not a problem in many other ecosystems. Transpiling, to name one example, is just not necessary. I think a sizable portion of multi-lingual devs would agree on 2 as well. But 3 is difficult to quantify. Is it a good thing that I can me a good chunk of money building React/Redux apps when I could've built a saner, more stable version in Django/Rails/Elixir in a fraction of the time if the only downside is that it requires a full page reload? I'm not so sure anymore. But it's worth a discussion. Point 1 and 2 strike me as obvious enough that bringing them up ad nauseum on HN is just pointless and frustrating.