Google stomping all over standards processes with shit no one needs. That's mostly in HTML-land, but it bleeds into JS land as well. Like the "requirement" for JS to have C++-like classes (it doesn't) that mesh really poorly with prototype-based inheritance and still causes lots of problems (have they solved auto-binding instance methods and private fields yet?).
No standard library, none. That is why you get so many libraries doing the same thing and people importing left-pad.
Abysmal rate of development. That is, however, the result of being developed in the open (somewhat) and needing consensus on new features. This is preferable to browsers and runtimes just doing whatever the hell they want.
NIH-syndrome. Mostly from the frontend crowd, but node.js-ers are bad, too. The audacity to think that no one ever in the history in the world made better UI systems, better backend systems etc. All while painfully re-implementing and re-discovering the last 30-40 years of computer science and engineering.
Beating a dead horse: weak dynamic typing is the worst typing, and Javascript is that. Strong dynamic typing I could accept.
For the most part, the class syntax in JS is just sugar over the top of Prototype based inheritance. It's not my preference, mostly in that it doesn't work well with too many layers of inheritance, and enhancement tends to work better imo. I'm also more in favor of a more functional/modular approach.
In terms of private members, you prefix # for that...
class MyClass {
#myValue = 6;
...
}
Plenty of languages don't have a "standard library" including, for example, rust... there are benefits and disadvantages to this. The benefit is you also aren't stuck with a poor interface for certain things. You can use Deno/jsr std if you prefer.
As for the rate of development... abysmal is not the word I'd use. JS in general has been prolific to say the least and nearly exponential in pratice.
In terms of reinventing UI... if it's just a retread, then why is it there hasn't been a better cross-platform UI toolkit developed? You can do more with HTML+CSS in terms of layout than any other UI native toolkit and even then, most fall short and are limited to only one or two platforms. That doesn't even capture accessibility needs and requirements.
Weak dynamic typing I can somewhat agree with. However, given the origin of JS was mostly for input validation, it makes a twisted bit of sense. Just looking at how falsy behavior works in JS makes a lot of sense, the only minor outlier is 0, but that being falsy in nature comes more from other languages than the subject area of JS itself. I do think that sometimes dealing with undefined, undefined as a value and null is a bit of a pain, and kind of wish that the concept of Option<T> was more prevalent when JS was created.
In terms of your favorites, I've used C# since before the first release and am very familiar with it. There's a lot to like. That said, it doesn't stop the environments that use it from turning it into a twisted, byzantine mess of layers of abstraction and indirection that destroy every bit of performance it's capable of in the name of "Enterprise" practices. I also like the C# language. I just hate most of the C# codebases I've had the displeasure to work in. It's far from exclusive to JS.
I basically agree with your corrections/details, so no need to drag the discussion out :)
Except this:
> you can do more with HTML+CSS in terms of layout than any other UI native toolkit
I cannot in my life understand how this worldview is so common. It's like no one has ever seen a native app before the primitive Electron and web apps that struggle to display a few lines of text.
Turbo Vision from mid-1990s was arguably a more powerful GUI toolkit than HTML+CSS. Most of the things HTML+CSS for which pat themselves on the back were available in Qt in early 2000s.
There are some niceties about some native UI frameworks... that said, they still tend to have shortcomings. Often don't work or have a really poor experience for cross platform dev or deployments. It also doesn't necessarily cover accessibility integration, which I've experienced the most pain in dealing with (at least going back into the 90's, far less experience recently).
Yes, out-of-the box cross-platform is the main thing going for HTML+CSS. It's an undeniable fact.
As long as you don't do anything fancy on the page, accessibility is now great for HTML, too.
However, since it lacks everything that makes a UI useful [1], all the custom components everyone is busy re-making from scratch make the web a miserable place from accessibility point of view [2]. Much worse than native frameworks which often enjoy deep integration with the underlying OS's accessibility features.
[1] See the work by https://open-ui.org and the many, many, many UI components most UI toolkits take for granted and have out of the box
Not at seemingly infinite scale, no. The days of troll farms where a single human operator working in a low wage country is commanding 24 phones and social media accounts at once in a purpose built workstation are probably numbered.
There's a whole section here about how to brute force the hashs. You don't even need the full hash... just a shortened version using the first few chars.
I'm dubious. Searching for globally unique commit IDs is still a least a million+ request operation. That's easy enough in a cryptographic sense but the attack in question requires banging a web UI, which is 100% for sure going to hit some abuse detector. I really don't think you can do this in practice, and the article certainly doesn't demonstrate it.
Honestly, I remember now getting this interview, but it isn't it. I looked through the whole transcript and the interviewer was not skeptical of computers, there was hardly any mention of the Dewey Decimal system etc.
Is there anything else? That one I had only audio, it a radio show I believe. And the guy was talking about music and dewey decimal system etc. Must have been 80s or 70s. It wasn't Alan Kay
