Hacker News Reader

Comment by jdonaldson | Hacker News Reader

I would argue jquery has been more instrumental to the success of the web than javascript. It was the stroke of pragmatic genius riding the back of a half-baked language ( a prototype-based language trying to ape the style/name of Java, solely as a cheap cash in of its popularity).

View on HN

AJAX and the Web 2.0 hype around 2005 are responsible for the success of the last few years. There were also YUI, prototype.js, ExtJS before anyone used JQuery. But thanks to the popularity of JQuery around 2010 we got at least two JS functions in HTML5 browsers. And the browsers got more compatible with the HTML spec. Before you had to write for IE, Firefox and Webkit with fallbacks for various bugs. Since around 2013 vanilla Javascript is enough, and JQuery popularity is declining. There are React/Angular/Ember/etc nowadays but hopefully soon ShadowDOM will be supported by most HTML5 browsers.

Just to clarify: YUI and ExtJS both came out after jQuery, jQuery came out in 2006, and jQuery is actually more popular now than it was a year ago (or in 2010 - see the figures cited in the blog post).

Not that this is wrong, but YUI was released less than one month later and had been in use at Yahoo for a few months.

http://yuiblog.com/blog/2006/02/13/the-yahoo-user-interface-...

Yeah, if you want "older", you need to reach for Dojo or Prototype (with Scriptaculous obviously).

Wikipedia lists YUI started in 2005, so several months before JQuery. And ExtJS afaik started with prototype.js support and later added JQuery.

jQuery is a dependency of Ember. Angular (at least 1.x) requires either jQuery or its own subset, jqLite. Cross browser incompatibility is still a thing and jQuery remains a very good tool for abstracting the differences.

I still use jQuery very heavily in projects not already built on frameworks. It's way easier for me to add a new module or function to a monolith like that by writing a little jQuery than trying to overhaul the whole project to use Angular/Ember/React/Vue/Mithril/whatever.

Thank god for jQuery for this exact reason. It's been a refreshing constant in the exhausting mess that is the JS framework world, and the fact is that most software is "monolithic."

The weak DOM API not JS was what plagued the early web period from 2000 - 2005.

Not weak, but every browser had their own non spec API functions to traverse the DOM in various ways. JQuery offered an easier cross browser API, whereas otherwise one would have to handle various bugs and different API functions.

Particularly events. I think there are technically 3 different ways to create event handlers, one for each of the major js engines of the early 2000's.

Listeners are complicated, but .onclick and friends work everywhere and always have.

For the simplest cases, yes, but even there you could easily find oddities with things like cancellation, bubbling, timing with regards to other events, support for various element types, etc.

Here's a whole list just for click, which is probably the most standardized event handler, and doesn't even go back before the IE5.5 era:

http://quirksmode.org/dom/events/click.html

to double down, touch events on mobile browsers are in a different order than mouse event in regards of onclicks, so that a touchend and mouseup fire at the opposite side of the click event on some mobile browser but not all.

Until you want to add more than one onclick event.

Maybe I'm just old, but from the perspective of the current day, I'd call the early web period 1990-1995. Or possibly 1990-2000.

2000-2005 is the early "dynamic html" period. Or at least that's what I put on my resume at the time.

Actually DHTML goes back to 1997 (with Netscape Navigator 4 and Internet Explorer 4).

I just had a pang of nostalgia after I checked and saw that Dynamic Drive is still online :)

Dynamic Drive was how I learned Javascript back in the day! Pulling apart how the author of a script had written it, to see how to get the snowflakes drifting down my page become leaves instead ;)

Not maybe, but yes.

1990-95 is pre-Netscape -- that's kind of pre-history.

(I too am old enough to remember the internet before the web. You really had to work hard to goof off at work when all you had was nntp, archie and gopher.)

What, you didn't use mosaic? I used mosaic, on VMS no less.

I did too, on SunOS. I remember the early days when renavigating to a page would just pull up the cache version and not refresh automatically -- I didn't understand why wsj.com wasn't publishing new content daily.

And I still miss the spinning globe.

I used Internet Explorer on SunOS for testing Applets on Windows and Unix.

> You really had to work hard to goof off at work when all you had was nntp, archie and gopher.

Nonsense, I could goof off quite easily with nntp :)

I don't think that's accurate (or fair, really). Jquery was one in a list of dozens of libraries that did similar things at that time, but ajax/interactivity had been part of the web mindset for years before jquery existed. Jquery just won the library war.

Yep, I'm surprised jQuery isn't older, as I do recall using dojo and prototype before 2006.

jQuery was a huge timesaver, as was dojo, mootools, prototype, etc and not having at least one of these libraries would have sucked. And although jQuery "won" it could have easily been any other javascript library.

Even if these libraries somehow didn't exist, there was tons of cross-browser snippets plastered across the web that dealt with most of the common problems (ajax, event listeners, etc).

What drew me to jQuery over Prototype in particular was the fact you could use css-style selectors to get and operate on several elements at once. This often reduced several lines of code in a loop to a couple chained methods.

Prototype first came out in Feb 2005. Dojo 0.1 was Aug 2005, and 1.0 (first stable, after a total rewrite) was Nov 2007.

I do remember using Prototype with script.aculo.us and thinking it was pretty neat to add some effects to my server-side-created web apps, but that they'd largely stay server side. The first time I saw jQuery, that was when I thought "Yeah, okay, we can actually start building some serious client-side interactions here" instead of just server-generated HTML pages.

jQuery took me from hating (but tolerating) javascript to actually enjoying client-side web development.

jQuery was important not just because of the feature set, but because it completely dominated the industry. jQuery and its ecosystem of plugins created a single community for JS developers. A lot of subsequent work wouldn't have been as successful without that strong community in place to spread ideas.

The other thing it did well was change the thinking about how JavaScript should interact with the DOM. In the past you'd have libraries that made DOM access easier, but they largely followed the same conventions as the underling APIs. jQuery broke from that and completely re-envisioned the API layer. The underlying APIs were treated as an implementation detail; this approach is something that subsequent libraries have adopted too.

Exactly. Method chaining was a novel feature back then

It won because it has a great API and excellent documentation.

jQuery is pretty much the only library or framework made 10+ years ago I can use and still feel like "this is actually well-designed". Frameworks generally obviate the need for it, but if a framework isn't an option and if I'm going to be writing more than 100 lines of code (otherwise I'd suck it up and use plain JS), I always still use jQuery.

I would give that distinction to XMLHttpRequest. It's why we could have Gmail, Google Maps, etc.

Not really. They used the forever frame, dynamic script tags with jsonp, dynamic <img> tags, and other tricks (at least at the beginning).

Reply on news.ycombinator.com