TC39 definitely needs a better, clearer website for people to browse then just using github repositories. So this is quite welcomed.
At the moment though there's nothing in it. No search capabilities or unique content and practically all links redirect to github. This isn't an MVP it's a prototype, with little value at the moment. Hopefully things will improve soon, but by shipping such a skinned down website they won't get the product launch effect again.
> Ecma TC39, the JavaScript Standards Committee, is proud to announce that we have shipped a website [0] for following updates to the JavaScript specification.
Seems to be the new website from where I'm sitting. The spec is linked from the header. It might be an old instance, but it is clearly meant to be used as part of the new.
It's a product, and they've just launched it. The fact that they aren't a for-profit company is not really relevant.
Nobody uses the product right now, and by product I mean the TS39 website. And assuming that one of the purpose of the website is to be used they'll want the site to have traffic and traction or it's just a waste of effort.
Product launch can have a very big effect on future traffic - that means that many people are exposed to the site, and if it proves useful they might also start using it. As it stands no one is going to return to this site instead of going directly to the github account.
"What search capabilities are you hoping for that are currently missing? "
At the top of the hat, searching for specs in different stages? or by topic, or date. Without search the site can't be a location for discovery, and since it has no new content, it isn't a consumption website either. If you look at it, its basically a glorified landing page. Not something that is useful for the majority of its audience, I believe.
A websites doesn't have to be a product. It can just be... an informational website. Not everything that goes on the internet needs to be hyperoptimized.
Never seen that logo before. For some reason they set the numbers in Roboto but “TC” in something Avenir like. [EDIT: It’s probably Neutra Text, which the JS logo uses]
Javascript is the first language which targets the DOM. Great first step ! The future of Javascript, i think, will target all platforms. Consider it as a meta language to compile to native platforms is one way for the adoption of JS itself.
>Consider it as a meta language to compile to native platforms is one way for the adoption of JS itself.
Compiling to javascript was never a good idea, it was only ever, at best, an adequate kludge for which no alternative existed. Now we have Webassembly, an actual bytecode which can run on native platforms and the web, and which doesn't carry the semantic issues that come with compiling one high level language to another poorly written, poorly designed, poorly typed one and pretending it's basically the same as machine code running on bare metal.
We can finally drive Javascript back to the web where it belongs and where it should remain quarantined.
To you. JavaScript and TypeScript are the languages I am most productive in. There are use cases that I would benefit from being able to compile JavaScript to another language, or native code.
Neither do strawman arguments. The point is not that everyone who uses JS is a script kiddie, which is indeed false. The point is that it's the self-reinforcing wide adoption that encourages further use, not the great language design. I would say the same of Windows and English, even though there are great English speakers and Windows devs.
Not particularly disagreeing with you on the language design bit, the language has many flaws indeed.
However I don't really agree on your interpretation over the reasons why the language is so popular. How are things like Node.js or supersets like Typescript/Flow self-reinforcing the language? If JS was a purely browser oriented language the argument would make sense as it's basically _the_ only language available (other than WASM, but that's still relatively niche), but it hasn't been for some time now.
Lots of people learn coding by toying with JS (because of ease of setting up an environtment) -> large base of users willing to use the JS version of anything -> lots of development time invested in patching over its problems -> "Stars above, JS is so fast and has such great libraries!"
None of that involves JS being a well-designed language, which was my original point and which you seem to agree with.
I mean, whatever your thoughts are on the merits of the language, it's hard to deny though that Javascript's monopoly of the browser fueled its rise in popularity and propelled it to its current state.
There are people who like the Javascript design. Compared with it's peers like Python & Ruby, javascript is more functional & "lispy". It's faster, and non-blocking I/O is the default. Closures are cleaner. Some people prefer prototypical inheritance compared to classical inheritance.
> There is a reason it is used so much outside of webpages.
How are you coming to that conclusion? The current Tiobe index still has Java, C and Python topping the popularity charts. [1] If you put any credence into whatever this W3Tech site is about, Javascript's popularity as a server-side language even seems to be overblown. [2]
Your comment is quite confusing. What is your argument? According to your references Javascript is the 7th or 8th most used programming language. That counts as ‘a lot’ to me.
> There is a reason it is used so much outside of webpages.
I'm asking how you are coming to this conclusion. I personally do not see any evidence to support this and if you take any of these popularity/usage surveys into account, they do not support this either.
> 7th or 8th most used programming language. That counts as ‘a lot’ to me.
I mean, come on...that's not the same thing as "is used so much outside of webpages".
Yeah, it isn't really used much outside of webpages and web tech like Electron. The only applications where it has non-trivial usage are those that cater to webdevelopers.
I think that these surveys are really unreliable but even if a language is only used in 1% of all projects that is still a lot of projects. That’s why I think 7th or 8th is still a lot.
Without the weird gimmicks, it's not Javascript. Many people have "fixed" Javascript in different ways, creating new languages in the process. One of these is even quite popular, aka Typescript.
The new syntax for #private vars is truly terrible and the arguments for it are not compelling enough - some flavor of soft private would be OK. Please, if anyone from TC39 is reading, take the criticism seriously.
Plenty of people have brought up the "looks ugly/terrible/dumb" criticism and they seem to be sticking with the "total private encapsulation" proposal.
Personally I don't have a problem with it, since Typescript offers "soft" access modifiers that work fine for any use cases I can think of.
40 comments
[ 2.8 ms ] story [ 109 ms ] threadAt the moment though there's nothing in it. No search capabilities or unique content and practically all links redirect to github. This isn't an MVP it's a prototype, with little value at the moment. Hopefully things will improve soon, but by shipping such a skinned down website they won't get the product launch effect again.
[0] https://tc39.github.io/ecma262/#menu-search-box
Seems to be the new website from where I'm sitting. The spec is linked from the header. It might be an old instance, but it is clearly meant to be used as part of the new.
[0] https://tc39.github.io/
https://tc39.github.io/ecma262 , which is the github pages for the ecma262 repository, has existed since September 2015 - https://web.archive.org/web/20150929233059/https://tc39.gith...
What search capabilities are you hoping for that are currently missing?
Nobody uses the product right now, and by product I mean the TS39 website. And assuming that one of the purpose of the website is to be used they'll want the site to have traffic and traction or it's just a waste of effort.
Product launch can have a very big effect on future traffic - that means that many people are exposed to the site, and if it proves useful they might also start using it. As it stands no one is going to return to this site instead of going directly to the github account.
"What search capabilities are you hoping for that are currently missing? "
At the top of the hat, searching for specs in different stages? or by topic, or date. Without search the site can't be a location for discovery, and since it has no new content, it isn't a consumption website either. If you look at it, its basically a glorified landing page. Not something that is useful for the majority of its audience, I believe.
It looks bonkers, and not in a good way.
Compiling to javascript was never a good idea, it was only ever, at best, an adequate kludge for which no alternative existed. Now we have Webassembly, an actual bytecode which can run on native platforms and the web, and which doesn't carry the semantic issues that come with compiling one high level language to another poorly written, poorly designed, poorly typed one and pretending it's basically the same as machine code running on bare metal.
We can finally drive Javascript back to the web where it belongs and where it should remain quarantined.
If you're talking about compiling from javascript to another language... that seems neither useful or desirable.
To you. JavaScript and TypeScript are the languages I am most productive in. There are use cases that I would benefit from being able to compile JavaScript to another language, or native code.
“There are better operating systems than Windows. There are also better languages than English.”
The language has its flaws of course, but come on. Gatekeeping doesn't do anything constructive.
However I don't really agree on your interpretation over the reasons why the language is so popular. How are things like Node.js or supersets like Typescript/Flow self-reinforcing the language? If JS was a purely browser oriented language the argument would make sense as it's basically _the_ only language available (other than WASM, but that's still relatively niche), but it hasn't been for some time now.
Lots of people learn coding by toying with JS (because of ease of setting up an environtment) -> large base of users willing to use the JS version of anything -> lots of development time invested in patching over its problems -> "Stars above, JS is so fast and has such great libraries!"
None of that involves JS being a well-designed language, which was my original point and which you seem to agree with.
How are you coming to that conclusion? The current Tiobe index still has Java, C and Python topping the popularity charts. [1] If you put any credence into whatever this W3Tech site is about, Javascript's popularity as a server-side language even seems to be overblown. [2]
[1] https://www.tiobe.com/tiobe-index/ [2] https://w3techs.com/technologies/overview/programming_langua...
> There is a reason it is used so much outside of webpages.
I'm asking how you are coming to this conclusion. I personally do not see any evidence to support this and if you take any of these popularity/usage surveys into account, they do not support this either.
> 7th or 8th most used programming language. That counts as ‘a lot’ to me.
I mean, come on...that's not the same thing as "is used so much outside of webpages".
Here’s a Node package:
https://www.npmjs.com/package/left-pad
Node.js is for server side programming, not Javascript in the DOM. The package has almost 3 million downloads, weekly. To me, that is a lot.
we have shipped a website
That is an unusual phrasing. Was it too simple to use "created"?
[0] https://github.com/tc39/tc39.github.io
Personally I don't have a problem with it, since Typescript offers "soft" access modifiers that work fine for any use cases I can think of.