Comments by "John Smith" (@JohnSmith-op7ls) on "Theo - t3․gg" channel.

Watching the web front end community reinvent the wheel every 6 years just to eventually end up with what they already had, declaring it too complex, then starting on the next mission to reinvent the wheel… always funny.
50
“Happily make physical hardware” 1. “Physical hardware” is redundant. 2. Sorry but this programmer is now an expert in software and designing AND making every kind of computer hardware? Lol. So they’re basically a one person Intel, Western Digital, Samsung Electronics, and SuperMicro, while also being the best programmer ever. This is just stupid. If you’re going to make up fantasy ideals, just say they’re an actual god who simply wills any problem out of existence, so they don’t even need a computer solution for it.
7
@Z4KIUS Trivial performance gains like this rarely matter to begin with. Spend your time addressing issues that cost real money or adding features that make it. Chasing tiny page load speeds is just mindless busywork.
7
Contrived performance examples are just clickbait. This stuff is useless in the real world. Simple sites like this won’t benefit from trivial speed increases and more complex ones aren’t represented by demos like this and will have the real bottlenecks in places you can’t address with this stuff.
6
@Vulume Not to this degree or this pace. I haven’t seen the same churn on desktop, game, or even mobile UI. Web tech sucks but rehashing the same frameworks and libraries over and over with new names and syntax differences doesn’t make it better. The root of the problem lies with the slow and inept W3C. Has since they formed.
4
@Vulume It’s far worse than desktop, web is still struggling to provide what desktop apps were doing decades ago. Doesn’t matter how common desktop app dev is now, totally besides the point. Replace it with native mobile if you want and it still holds. Point is, web front end is held back by the incompetent W3C and people waste their time and energy on a treadmill trying to rehash more ways to do what desktop and native mobile apps have been doing for ages with far less hassle. HTML was never meant to do what it’s asked to, CSS has always been a hack to get it there, JS is awful, the APIs lag decades behind where they should be. I mean, HTTP didn’t even have a PUT verb until really late in the game. Imagine going two decades before the people making standards realized updates are a thing, and also they should be called PUT because being unintuitive is fun or something.
4
Yeah, no. Not even close on that second part.
3
⁠ @fischi9129 Give the lame “skill issues” non-argument a rest. You’re comparing an app written in C to JS? That’s your big argument that over complicated JS frameworks are fine, because more complex stuff exists? Lol. That’s like saying you shouldn’t complain if you’re robbed because hey, others get killed! Being lazy has nothing to do with not wanting to suffer over complicated designs. Lazy is mot wanting to find a better, simpler way and just use whatever junk is already there, even if it wastes lots of time I’m the long run.
3
It’s pretty rare to be doing long running processes in a web front end which doesn’t need to be constantly manipulating the DOM. Doesn’t seem worth it to run components in their own threads by default when the overhead won’t give you any gains in most cases. Running code in a worker will let it run on its own thread, and you can get the number of logical cores in JS, but once it needs to do something to the DOM, you’re back in the main thread/process, and those workers are competing. So you’ve still got the UI bottleneck but now you’ve added the overhead of the web workers and managing them.
2
How? Because the W3C is slow and incompetent and has been since day 1. It’s why it took like 12 years to finish the HTML 4 standards and Microsoft and Netscape just did their own thing because people needed to get things done and couldn’t wait a decade for simple functionality to get standardized. This is primarily what kept to so much browser feature support fragmentation for so long.
2
@shapelessed It’s because the W3C is awful and operates at a glacial pace. So it took decades to make basic UI standards, no excuse for that. Most of what these frameworks do could and should be baked into the browser as APIs, and that’s the direction it will go, just that it’ll happen inexcusably slow. But eventually these frameworks will be as irrelevant as using tables for layouts (html emails aside)
2
@federico.r.figueredo Lol, what? How does using something inside an OS make it an OS? Nobody that knows what an OS is would say that.
2
@federico.r.figueredo Wait, you’re arguing that you can call a giraffe a hippopotamus but that doesn’t change the fact that it’s a giraffe, then you say it doesn’t matter what React actually is because the developers call it a library? Lol, you just argued against yourself
2
This is way older than intercooler. We were sending down html to dynamically update the page even before xmlhttppost was a thing. MS had Remote Scripting that was way ahead of its time and you could use hidden iframes to get the job done too. When xmlhttppost came out and everyone was failing with passing down bloated xml documents, parsing them, basically a slower version of what people do now with json, those of us who were smart saw through the AJAX hype and just passed down html chunks. Amazing it took this long for others to catch up and make intercooler/htmx and Blazor. Although they at least iterated a bit by letting you use web sockets. Buts neither is anywhere close to mainstream and most people are still throwing around lots of json and bloated clients to deal with it.
2
Most web apps don’t need a front end framework. Some libraries to make DOM access easier and handle UI stuff like modals snd tool tips if you really need them, sure, but a full on framework? We wasted so much time and resources on this fat web client side quest only to end up going back to server side rendering. Funny but sad, and expected.
2
Unless the team is allowed to make hiring and firing decisions, that’s a profoundly stupid policy
2
That’s what happens when you have an inept standards body that takes decades to roll out basic things. Everyone scrambles to add those features their own way and you get a hundred competing options. And of most of those will reuse one of the hundred competing options for various lower level functionality and so on. It’s why a hello world Vue app is what, over 1000 packages, all of which are going to be updated, abandoned, and wind up with security issues at random times.
1
People still use Wikipedia? Why? It’s basically Reddit with different structure
1
You guys have a circle jerk going?
1
Because everyone needs to make some amazing new framework to climb the ranks at work, or as a resume booster to land jobs, or maybe they’re hoping it’ll get big and they can make a business around it. None of these reasons have anything to do with making apps easier to build or maintain.
1
@shapelessed Yeah the W3C is awful. The HTML 4 standard took, what, like 12-14 years? Back then everyone gave MS flack for not following standards but the standards massively outmoded and MS had lots of enterprise customers who needed to get things done now. I just find it amazing that we can throw endless resources as a never ending list of frameworks and libraries and learning to use them, updating tooling, dealing with quirks, but we can’t put the resources into faster development of web standards and just bake most of this stuff into browser APIs where they belong
1
@fsbgaming1588 Yeah feature bloat is a problem at all levels of development. It’s always a balance that takes long term experience to maintain. For non-trivial apps, a framework is usually a good idea but you also have to be willing to stick with that framework long enough to provide significant benefits over the cost of learning it, getting good at it, learning related tooling, dealing with ongoing updates to the framework and the regression testing test results, dealing with older versions of the framework (do you periodically migrate to the latest? How often? What’s the cost there?). Most devs just focus on how fast you can build something. Many don’t think past dev time and effort to do enhancements. There’s more to app lifecycles than just build and enhance. People will say they understand the full scope of what maintenance entails but not many seem to in practice. Companies like Google make this worse as they usually have one team build something, then move on to a new product while another team comes in and has to deal with maintenance. It’s one reason why so many Google products launch, then never see tangible enhancement or bug fixes.
1
@ But this isn’t about addressing relevant performance issues, it’s about pointlessly squeezing out a bit more, in a contrived demo, just for the sake of it.
1
@Evan-dh5oq Businesses can’t make demos? Making a page load 50ms faster isn’t making this company any money, or saving it any when you consider the effort spent on implanting all these tactics just to see how fast they could make the site. That effort could have been better spent on adding functionality that makes or saves money. Or if there’s literally nothing else to develop to that end, they’re probably overstaffed.
1
@ It’s clear you have no idea what you’re talking about
1