I'm going to pick on Javascript as an example, I'm sure there are examples in other areas.
Dynamically loading content in web applications has been supported for almost 15 years [0]. Manipulating the DOM has existed longer than that, remember the browser wars [1]? How many JS frameworks have been invented in that time? How many of those caused a fundamental positive shift in the output vs. just a slightly different way to arrange the symbols and words in a text file?
The concept for a tablet computing device happened in the 1960's at PARC [2].
Maybe reinventing the wheel every few years is holding back our ability to make true leaps forward.
[0] XHR support in IE7 in 2006: https://en.wikipedia.org/wiki/XMLHttpRequest
Both vision and basic research are basically neglected to the point it's laughable. Instead, what gets funded are half-baked ideas turned into MVPs by Ivy League CS grads with silver spoons hanging out of their mouths, whom unironically wear company hoodies (they didn't get the memo Mike Judge was making fun of them), and who hopped on the latest bandwagon for whatever trend, burning millions creating an acquisition target that adds no value to the world, at best.
Oh, and a cool 90% of the modern web development ecosystem probably shouldn't exist and is pretty much poorly-engineered trash.
Ads encourage people to buy stuff they probably don’t need. They incentivize clickbait and tracking. And most online advertising is probably a waste of money anyways. [1]
How then, you might ask, will content and services be paid for?
Sponsorships are not a solution for the same reasons.
The Brave browser is not a solution. [2]
I think the solution is to pay for everything with donations and crowdfunding. This model incentivizes doing what’s beat for the end-user. Great examples of this include Wikipedia and the Internet Archive.
If ads were banned, many websites would die, and I would be ok with that. Only the ones that are the best for users would survive.
[1] https://thecorrespondent.com/100/the-new-dot-com-bubble-is-h...
Good will doesn't scale up enough to cover the entirety of the internet. And ultimately that leads to more paywalls being put up.
For example, some simple leetcode question that will filter out a fair number of people. After that it should just be based on first come first serve.
Why? Becauae as you keep increasing leetcode difficulty, you're going from one extreme (can't code fizzbuzz) to another extreme (competitive programmer or someone who just got really lucky).
Most will agree that there's lots of luck involved in the interview process so why not have most of this luck simply be a matter of applying on time?
By having the luck factor in leetcode, we have to keep those skills sharp for the rest of our careers.
This is also a loss for the employer since the leetcode prep is a significant time sink which has nothing to do with the job.
If when you applied was the factor used to break ties, recruiting cycles would also be shorter and candidates would bring more job relevant skills.
Overall it's a win for everyone but it won't happen because the people who could actually change this have no incentive to do so.
"Let's grow!" is the mantra in the tech industry, which to me is parallel to the idea " let's keep augmenting the population of this planet, there are resources for everyone!"
A lot of people may be nostalgic for the 90s or what not, but it's a good thing the world moved on, standards evolved and tech got better.
Back in the day you walked to the phone booth once a week to talk to family members that are far away. Nowadays they can contact you 24/7.
- Online advertising is a great business model, and targeted advertising doesn’t harm anyone.
- Technology doesn’t threaten privacy, it enables it.
Also, making MULTIPLE products that EVERYONE uses is really, really hard. These companies manage to do that and have enough budget to experiment.
Also Also, getting a software engineering job in a FAANG is still, 20 years later, a solid path to prosperity, which is rare
JavaScript moving server-side is a massive regression in software engineering
Chrome is the new Internet Explorer. That is not a good thing.
Containers are another "worse is better" situation and it's shocking that docker has won so hard given its relatively flawed implementation on OSX.
Rust is overused.
Go is too low-level for its ultimate sweet spot as a language for building command line tools and services.
There is no really great build tool for JVM languages, yet I still find their build ecosystems far more robust than those of JS or Python.
These startups have the same goals as cancer, growth at all costs.
Vue is going to beat out React and Angular.
Anti-patterns are just patterns that some people don't like.
Actually... you don’t need to replace your tool X every week just because a new tool Y has been invented that does the same thing but has a much less support.
- they do not represent the majority of tech companies
- the majority of devs out there do not work for them/won't work for them
- their techniques/processes/rules do not apply to the majority of tech companies
Ex user crested add an event, updated add an event.