View Single Post
chucker
 
Join Date: May 2004
Location: near Bremen, Germany
Send a message via ICQ to chucker Send a message via AIM to chucker Send a message via MSN to chucker Send a message via Yahoo to chucker Send a message via Skype™ to chucker 
2019-06-14, 14:01

We all know the tale: around 1994, Marc Andreessen and friends figure out that the World Wide Web on the Information Superhighway can be more than a bunch of hyperlinked academic documents on university computers. By adding HTML elements that seem basic and obvious these days, like <img> and <table>, they make web pages a lot more exciting for every day users to look at.

They package it up in an app they call Mosaic… then they get in a fight, rage quit, and re-brand as Netscape Navigator. It sells as shareware for $49 and more, but nobody actually buys it that way, because Netscape smartly makes deals with ISPs to bundle it with your home Internet, and with magazines to put it on a CD. Yes, those were the days when your magazine came with a CD glued to it. Also, the days when you read a magazine.

It was a huge hit.

And then Microsoft noticed. Windows 95 was mostly a successful launch, but if there's one blind spot they overlooked, it was the coming boom of the Internet. They needed an Internet Explorer®. So they did what everyone unscrupulous would do: license Mosaic with a deal where Spyglass, its new owner, would get a percentage of the revenues. Then give the browser away for free.

It turns out when you have a widespread operating system and bundle another piece of software for free, competing vendors have a hard time, well, competing. By the time both Netscape and IE reached version 4, the tides turned, and Microsoft had the popular and arguably superior browser. It had CSS at a time when Netscape bet on the ill-fated JSSS. And, more importantly, it was already on your computer. Even on the Mac. Why install something else?

Then IE reaches version 6 (on the Mac, it never quite did). Here's where things get weird. We're at around 2002 now. Netscape, since acquired by then-media behemoth AOL Time Warner, is scrambling to get any working new version to ship at all, as they had decided to open-source their code base, and then open source community had in turn promptly decided to rewrite the entire thing. Losing half a decade to an aggressive competitor may not have been the best of strategies. Netscape 6 does eventually ship, but few notice.

Microsoft is working on its next major OS release, codenamed Longhorn (eventually known as Vista). With their browser having captured almost the entire market, they don't see the point in evolving it much further, so they significantly scale back its development team.

Bad move.

People may not have cared about Netscape 6, but they do take notice when a never-heard-of project called Mozilla takes out a full-page ad in the New York Times. Firefox 1.0 is here. Like phoenix (one of its previous codename) from the ashes, Netscape's remains were reborn in the Mozilla (originally "Mosaic killer") organization, and they figured out plenty of missteps along the way. Turns out Netscape Communicator 4, which contained a browser, a mail client, a kitchen sink and a yacht, was kind of a poor design, and people really want a lean and mean browser. Firefox is exactly that.

Mozilla is also lucky in that IE has a problem: in its Windows XP variant, it grows a reputation for massive security issues, and its coming Longhorn/Vista fashion, with an improved security architecture, won't ship for another three years.

So Microsoft, rightfully, got slapped back awake. Firefox grew popular as an underdog, and Microsoft reacted by once again forming a proper IE team that would ship versions 7 and beyond.

It is now roughly 2004.

I'm gonna leave out Opera here, and will skip some Safari history, except for the crucial bits.

For about half a decade, IE had shipped as the default browser on the Mac. It wasn't a terrible browser, and its engine was actually significantly different — better ways of standards compliance, but worse in some other respects — than that on Windows. But it wasn't a path for the future for Apple. Instead, they shipped Safari. Along the way, they hired away some folks from Netscape, which made for popcorn-worthy discussions.

It is now 2005.

At this time, Microsoft is still a bit slow to react. And there's another problem. I started out this article by poking fun at how academic early HTML felt. It had elements like blockquote and cite, but none like img. Well, leaving aside cause and effect, the process of evolving HTML continued in this kind of culture. Or at least, that's how Apple, Google, Mozilla and Opera felt.

They were annoyed by how the W3C, the standards committee tasked with moving HTML forward, was focusing on features that didn't in their opinion meet the mass market's demand. So they formed their own committee, the WHAT WG.

Perhaps in a similar vein to how the early Netscape days gave us images, tables, and scripts, the WHAT WG gave us video, audio, canvas, a much expanded array of form controls like sliders and date/time pickers, and more. They also pragmatically looked at Microsoft's weird XMLHttpRequest spec, took what was great about it (hint: it's not the XML part), branded it AJAX, got Microsoft to agree with those changes, and even eventually worked to supersede it with the Fetch API.

They moved the web forward. Not necessarily always in the right direction, but mostly in a good one.

At the time, those four founding members collaborated and contributed plenty. For example, <canvas> is basically Apple's; it was also used in Dashboard (R.I.P.).

Today is no longer that time. Times change, markets change, and motivations change. Each one differently. Let's fast forward to 2018.

Actually, hang on. I mentioned Google above in the WHAT WG. When that was founded, Google didn't actually have their own browser to speak of. That came a few years later. How was it formed? By forking WebKit, the project underneath Safari. So in its early days and for quite a few years, Apple and Google weren't just working together on specs in the WHAT WG; they were also working together on concrete implementations in the WebKit project.

But, again, today is no longer that time.

Today, Apple doesn't seem quite as interested any more in contributing this much to the web. They still do pitch ideas, and Safari continues to bring its own take to the web, but it's not the leading force.

Today's Opera is basically done. They were always a niche for end users, but played a big role in the early evolution of CSS.

Microsoft, in 2014, decided to revamp their rendering engine to modernize it, but threw in the towel by 2018. IE was basically dead, and for a variety of reasons, its successor Edge didn't make much of a dent — despite, again, being the default shipping browser in Windows. One of those reasons: Windows itself isn't actually that big a deal to Microsoft any more. They have different priorities now. A piece of software on the desktop almost seems quaint to them.

Mozilla's heyday seems in the past. When Google Chrome appeared, it took market share not only from IE, but also from Firefox. (And from Safari.)

And that brings us to Google.

You see, Google took Microsoft's playbook from the 1990s, and doubled down on the evil genius. Microsoft leveraged their near-monopoly in operating systems as a starting point to get IE to spread widely. Google took their existing near-monopolies in a search engine and a video platform to urge you to install Chrome. Microsoft encouraged their own technologies, and discouraged others', so their browser would make for the "best" experience. Google does this, too.

But Google gives it a brilliant twist. They're actually perceived as the good, cool guys while doing so. Web developers love Chrome. It has Web Driver and Web Assembly and Web USB and Web Bluetooth (amazingly, none of these are made up), and hey, Safari doesn't have some of them! Ergo, Safari is terrible, web developers say. In fact, Safari is the new IE, someone literally said.

Here's the insight Google had: all you gotta do is be "open" about your process. Make (the majority of) your product open source, and that's cool. Give your new features a published spec, and that's no longer perceived as proprietary. Is the spec still a draft? Have other browser vendors had a chance to comment on it? Have their concerns been heard? Have they had a realistic chance to implement the feature on their own? None of that matters. Web developers gonna web develop, and the feature already works in Chrome. If it works in Chrome, it might as well be the standard. If users don't use Chrome, that's on them.

"Buddy, why not get with the times and switch to Chrome?"

It's the new Optimized for Netscape Navigator™.

And now, we have another nail in the coffin: the "Living Standard".

The W3C and WHAT WG have dropped all pretense, and officially embraced the new reality: in the future, there really won't be an HTML standard at all. All there will be is vendors (mostly a single one of them) unilaterally publishing specs, having already implemented them in private.

We used to call that proprietary and evil, but now it's open and therefore good.

We know Microsoft won't be our ironic savior from Google — they've all but capitulated, and Edge is now Chromium-derived. We know Apple won't be either; they're still fighting their fight against web apps.

Our best hope is for another Firefox moment.
  quote