Hacker News Reader

Comment by tptacek | Hacker News Reader

This is the result of a standards process running in reverse: first the specifications, then the implementation. There was a time when the "Internet" could if you squinted right be defined as "the collection of people who make fun of that process and do the opposite"; their motto was (literally) "loose consensus and working code".

First you get things working. Then you standardize on them.

Complain when the Webkit team refuses to implement standards. Have they shown any unwillingness to support W3C standards once published? No? Then what's the gripe? The browser vendors are doing exactly what they're supposed to do: come up with new things. If the W3C was doing its job, it would watch those new things carefully and, when enough momentum existed, start ratifying them as standards.

It's not enough to say, "you could have used the same argument to defend Internet Explorer". To indict someone for "embrace & extend", you need both halves of the behavior: creation of new features, and unwillingness to support subsequent standards for those features.

View on HN

"Have they shown any unwillingness to support W3C standards once published? No?"

Unfortunately for your argument, the answer is indeed no. Apple is unwilling to support standards for touch events, due to its patent war against Android.

Sources:

[1] http://9to5mac.com/2011/12/19/apple-submits-invalid-patents-...

[2] http://www.webmonkey.com/2011/12/is-apple-using-patents-to-h...

Did the W3C standardize touch events, and Apple prevent Webkit developers from implementing the standard?

They showed an "unwillingness to support subsequent standards for those features", by blocking the creation of a standard in the first place. Stop being obtuse.

I think yours is the obtuse argument. The W3C has proceeded at a glacial pace on everything, not simply touch events. Is it your argument that Apple is preventing the W3C from standardizing corner radiuses?

Either way, I can hold two thoughts in my head at the same time: that Apple is indeed retarding standardization of touch events, while the W3C is retarding... everything else --- and then complaining about it.

Apple is a member of the W3C, so talking about Apple retarding the W3C process is not really in opposition to the idea that the W3C is responsible.

I do think it is fair to say that Apple is a key player in the W3C's issues, along with Microsoft.

It's a situation very much like the UN, where people opposed to the idea of cooperation use their lack of cooperation as proof of the idea that collaborative action is infeasible. Apple really wants people to use their walled garden as much as possible, so they are definitely among the parties intentionally sabotaging the W3C process.

Isn't the "walled garden" we're talking about here one of the most successful open source projects ever?

No. Apple maintains a fork of Webkit, which is the version present on all their devices. The browser in their walled garden is the non open-source browser they force you to use on their devices.

Contrary to character, they have allowed other browsers on iOS: http://itunes.apple.com/us/app/opera-mini-web-browser/id3637...

That's not a browser, that's a psuedo-browser. All rendering and javascript is done on the server.

I'm actually talking about iOS, especially the Apple App Store. Apple is threating to assert touch patents to prevent other vendors from implementing rich touch interfaces in the browser.

Eight years ago one might have called Java one of the most successful open source projects ever. Webkit could easily end up in a similar position. Given Apple's heavy litigation of their opponents, it seems almost inevitable.

So, the game has barely even begun yet. I suggest we revisit in 10 years and see how successful Webkit has been, for Apple, and our broader software community. (Because simply saying it is a success definitely suggests different things to different groups.)

You just totally changed what he said. He asked "Have they [the WebKit team] shown any unwillingness to support W3C standards once published?" Let's review the facts of the case you're talking about:

1. The WebKit team wasn't involved

2. Nobody was failing to support a published standard

Was somebody pulling a dick move in the article you posted? Yes. But it was a different group pulling a different dick move than the one tptacek asked about. If you want to argue that Apple's lawyers are a disagreeable lot, I suppose that's a valid point you could make, but this is not the thread for it.

blocking it? They just submitted their standards by the deadline, right? That's why you propose deadlines.... if planning was made assuming everyone would submit their stuff before said deadlines, well, maybe they need a refresher on project management?

> The browser vendors are doing exactly what they're supposed to do: come up with new things. If the W3C was doing its job, it would watch those new things carefully and, when enough momentum existed, start ratifying them as standards.

Both the browser vendors and the W3C could be doing a better job. But neither is really the problem here.

The problem is that WebKit, a single implementation, is utterly dominant in the mobile space (except for Opera on low-end phones in the far east). This isn't a criticism of WebKit. Any single implementation that becomes so dominant is bad.

Any single implementation, once it has vendor prefixed stuff in release builds, will lead to the problems the mobile web has now: People will use the vendor prefixes, after which that implementation can't remove them without breaking websites. And after that, other vendors will be forced to implement those prefixes in order to render the web in a non-broken way.

One way to avoid that is for vendors to not vendor prefix at all in release builds, but all browsers do that. If that is out, the only thing that can save us is for no single implementation to become dominant. But WebKit won on mobile - Android and iOS are by far the leaders, and both use WebKit; in fact, even most of the runners up do (Blackberry, Bada, WebOS, etc.).

My only criticism of WebKit is that, when it saw itself becoming so dominant, it should have been much more careful with vendor prefixing, preferably stopping doing so entirely in release builds. It's ok to vendor prefix when you have a competitive market, but if you do so when you are dominant or about to be, you end up with the current debacle.

In case people aren't aware of just how badly competing browsers are locked out of the mobile web, imagine a world where most of the buttons in Google Maps are broken in Firefox. Where bing.com doesn't work in Opera. Where non-WebKit browsers get a degraded experience in Google web search, and a barely-functional version of Gmail.

All of this is true on mobile phones today. And it's getting worse over time rather than better. If you want to build a browser that provides decent compatibility with popular mobile websites, you have little choice but to emulate WebKit's experimental (and often unspecified) features.

[Disclosure: I work for Mozilla on Firefox for Android, and as the editor of the W3C Touch Events spec.]

The web - the browser programmers, the HTML developers, certainly all levels of management - is incapable of learning from its mistakes - we've been over this with Netscape, then IE6, and now the exact same thing is happening with Webkit.

In my view what we've learned is the standardization process doesn't work. We had IE and Netscape doing what they liked and we got innovation; we got javascript, we got iframes, we got embedded video. Sure, each of these is a mess, to a lesser or greater extent. But they let us build the sites we want. Then in ten years of W3C we got nothing - just the blind alleys of xhtml and css 2.1. The web has only started moving again after the WhatWG basically said to the W3C "we're going to implement this stuff, you can either call it html5 or become irrelevant". We have learned from our mistakes. We're returning to the netscape vs IE6 days because they were better for innovation than the ten years of emptiness that followed.

The ten years of emptiness were caused by the browser wars and its ultimate, unequivocal victor when it, and the technology it supported, became entrenched for an extended period of time. It's not as though the IE team wasn't innovating because W3C stopped them.

During the emptiness, technologies like alpha-transparent PNGs started out but never got anywhere because the gorilla didn't support it and what were you going to do? We got this idea to create layouts without using tables, but oops, box model bugs, I sure was glad to be working around the legacy of the days of innovation, and hey, wouldn't position:fixed be great? You had upstart browsers that couldn't access many websites because in the happy days of innovation the web decided that if (document.all) elseif (document.layers) was a decent way of writing code and it's not like anyone was going to rewrite those. Remember that innovative native client technology a whole bunch of banks and DRM sites decided could be good for secure, controlled internet experience? I think they called it ActiveX, sure did wonders for usability and practicality of actually innovating browsers that could not support it.

You say we've learned from our mistakes but to me the use of -webkit- and continued use of user agent sniffing (buggy, natch) suggests otherwise. Who'll be updating those sites two years from now, when Firefox Mobile's or X Mobile's rendering engine is the innovative stuff all developers love?

It constantly amazes me how many people do not understand this. So many web standards advocates will cry foul the moment any vendor tries something out that has not already been standardized. Even Brendan Eich cried foul about Dart even though JavaScript was out in the wild for a year before Netscape even submitted it to a standards body.

Committees are not the right process for innovation. 10 years of overly complicated and unneeded standards from W3C should have taught us this.

Brendan Eich cried foul about Dart because other browser manufacturers, and web developers themselves, said they didn't want it, and Google did it anyway.

I was unaware that a poll had been taken of all web developers asking whether they wanted Dart. If developers really don't want it, they'll vote with their feet, so what's the big issue?

Also:

  > The big issue I have with Dart, which you
  > seem to consider inconsequential, is whether
  > Google forks the web developer community,
  > not just its own paid developers, with Dart,
  > and thereby fragments web content.
  >
  > "A Dart to JS compiler will never be "decent"
  > compared to having the Dart VM in the browser.
  > Yet I guarantee you that Apple and Microsoft
  > (and Opera and Mozilla, but the first two are
  > enough) will never embed the Dart VM.
  >
  > "So "Works best in Chrome" and even "Works only
  > in Chrome" are new norms promulgated intentionally
  > by Google. We see more of this fragmentation
  > every day. As a user of Chrome and Firefox (and
  > Safari), I find it painful to experience, never
  > mind the political bad taste.
  >
  > Ok, counter-arguments. What's wrong with playing
  > hardball to advance the web, you say? As my blog
  > tries to explain, the standards process requires
  > good social relations and philosophical balance
  > among the participating competitors."

--Brendan Eich, http://news.ycombinator.com/item?id=2982949

Basically he's arguing that the browser vendors of today (not developers/users) are the gatekeepers of what new technologies can be tried. It's like making browser innovation a UN Security Council where every member has a veto over doing anything. If you value progress and innovation, I do not see how you can support such a position.

If Google introduces something like Dart or NaCl, is willing to support standardization of it and even release a free implementation, but other browser vendors don't want to integrate it because they don't like it, that's their problem, not Google's. If that means popular apps become "Works best in Chrome," that is a self-inflicted wound.

"Basically he's arguing that the browser vendors of today (not developers/users) are the gatekeepers of what new technologies can be tried."

Anyone's free to try things. But those things will never become a part of the web platform if other vendors don't think they will benefit from implementing them. The people who make the browsers are gatekeepers of what ends up in the browsers. That's not a statement about how things should or shouldn't be; it's just a tautology.

The job of anyone who wants to create a web standard is to accept that fact, not ignore it. Ignoring it leads to single-browser "standards", which may get adopted by developers but will not become part of the open, interoperable web platform.

What are the things Webkit has "proposed" by releasing builds that you don't think should be part of the web platform?

And, what are the things Webkit has built that were "overridden" by the standards process? Where did Webkit actually get things wrong?

Just what comes to your mind.

First of all, I don't want to come off as argumentative. I think we agree on the fundamentals: Implementation should come before standardization. Vendors should feel free to experiment. The W3C should focus more on quickly standardizing things that are already working in the field.

Today, experimental and prefixed features used in the wild are creating real problems for interoperability and competition. But I don't think there's a villain here, and I'm not blaming WebKit. Both the browser vendors and the W3C are using processes that were designed to certain solve problems. Now those processes are creating new problems and might need to change in response.

But, to answer your direct question (although I think this is getting on a tangent from the original post):

> "What are the things Webkit has "proposed" by releasing builds that you don't think should be part of the web platform?"

Dart is an example of something that Google has proposed (not "WebKit" as a group) that I and many others - even others like Oliver Hunt within the WebKit community - do not think benefits the web platform. Some people would put Web SQL Database in this category; I don't have an informed opinion about that one. (Though I do know it's the reasons some versions of Gmail wouldn't even load in mobile Firefox.)

> "Where did Webkit actually get things wrong?"

Apple's "meta viewport" tag is something that they released in Safari and evangelized to developers without specifying it or getting community feedback. I had the pleasure of reverse-engineering it for Gecko. Among other things, it's not even implemented in a layer where it makes sense. (It should be part of style, not markup.) Apple never proposed it as a standard; other people eventually did and are trying to fix some of the problems. Unfortunately, the basic design is locked in by legacy content (and has been since soon after Apple shipped this feature in the first release of mobile Safari).

But frankly, I don't care so much whether they get things "right" or "wrong." As bad as meta-viewport is, at least Firefox and Opera and Android and IE can implement support for it without controversy, and be compatible with sites that use it. The way that prefixed CSS properties are (a) released to market by browser vendors, (b) evangelized to web developers, and (c) deployed on production web sites is creating a situation where new browsers are either locked out of existing sites or have to implement each others' prefixes. This is not a stable equilibrium.

This is an awesome comment, and thanks for writing it.

One thing I didn't have going into this thread that I have now is appreciation for how different this problem is in the mobile space than on the "desktop" web. It's easier to be frustrated with web standards in the desktop world, where Webkit seems like an unalloyed force for good. I can see how much trickier it is for your team given the fact that Gecko has been relegated to second class status by the market.

What doesn't benefit the Web platform is being tossed aside because of superior native development frameworks. That's a far bigger danger than someone forgetting to add a '-moz-' prefixed CSS properly.

A number of people seem to have a somewhat extremely optimistic view that they're going to make Javascript competitive in performance and memory use compared to native on mobile platforms. There's just no evidence that it's going to happen anytime soon. It's just not even close, despite how magical emscripten or Mandreel might seem.

With more and more browsing activity shifting to mobile, and with the majority apps being entertainment/multimedia/games, holding steadfast to the idea that there are not benefits to alternate VMs is more of a dangerous than transient breakage of pages.

Yes, the prefixes are a problem, but the Web needs to evolve fast to keep up. A period of incompatibility and fragmentation can be tolerated. It's happened before, chaos and disequilibrium followed by periods relative calm.

I'm more worried about stagnation, and a future where the web gives way more to the client-server internet days, with everyone running OS specific native apps that read and store data behind walled-garden non-HTTP clouds.

WebKit CSS gradient syntax was pretty explicitly overridden by the standards process in favor of a better one. Thankfully, WebKit changed its syntax to match.

Note that in a world where the Web was defined as "whatever WebKit renders", which lots of HN readers seem to be in favor of these days, we would of course be stuck with the old gradient syntax forever.

Do you trust the standardization process to make good choices? We got the stupid now-standard CSS box model over IE's sensible one (and now, twelve years(?) later, there's finally an experimental way to use the sane definition of the width of an item). We got IE's <object> with its clsid nonsense over netscape's <embed>. We got the wrong choice for iframes as well. I'm happy to believe the new gradient syntax is better (though if it's really that much better, surely webkit could have changed it themselves), but taken as a whole I'd sooner trust webkit's decision-making than the W3C's.

I think this is a valid concern. It is very similar to ActiveX with Internet Explorer, and, with the browser share Chrome is gathering, it isn't unreasonable to see people focusing on that.

--BUT--

With ActiveX, there was no other way to accomplish the same task. With Dart versus JavaScript, I can't accomplish anything in Dart that I can't in JavaScript. And, even if I decided to use Dart for some reason, it would be idiotic of me not to provide a JS version for non-Chrome browsers.

What would be really nice (heck, even something a standards body should have done) is coming up with a standard byte-code for browsers, so I can use any language I want. JavaScript isn't good enough, because I want my sites to load faster, and parsing the code shouldn't have to happen every time somebody visits a page.

> What would be really nice [...] is coming up with a standard byte-code for browsers

I used to think that too, until I read convincing arguments against it like this: http://www.2ality.com/2012/01/bytecode-myth.html

However, Javascript as the intermediate language does not follow. It may be true that a source-level intermediate representation has benefits, but the idea that the best intermediate representation is JS doesn't follow.

For example, late-binding, boxing/heap-numbers, etc all serve to make it extremely difficult to maximize startup performance, or heap overhead. It also makes the VMs much more complex to deal with optimizing such languages. TypedArrays don't really fix the problem, they just introduce others. For example, by forcing your to implement your own manual memory management everywhere. This is somewhat problematic if you're compiling a language with GC into TypedArray heavy code, since there's no finalizers in JS or Reference queues, you can't really do automatic release if a regular heap object holds into pseudo-objects in manually allocated memory.

The problem is particularly acute on mobile devices which are very resource constrained.

> I was unaware that a poll had been taken of all web developers asking whether they wanted Dart.

A few hours ago at JQuery UK someone asked Paul Irish what he thought of Dart. Then the whole room laughed. PI mentioned 'its complicated' and then that 'Dart has a place' and then everyone laughed again.

So a knee jerk unintelligent response by some random group of Javascript programmers, given without context, is supposed to mean something?

Paul Irish works with the Chrome team at Google.

Who said anything about Paul Irish?

The knee jerk reaction I was referring to was the "the whole room" of js guys that "laughed".

I think you missed the point of the comment.

Maybe. It sounds like the made fun of PI and/or Dart.

But I cannot understand either the point or why. If you could chip in with some context, I'd be grateful.

Haberman's post mentioned he was unaware of a poll of all web developers on Dart. I mentioned a recent example of such a thing with a large sample size. That's all.

It wasn't "any member having a veto over doing anything". Don't attack a straw man. It was every other browser manufacturer unanimously saying Dart is a bad idea, and specifically outlining reasons for doing so. The only possible position you could have in support of Google here is that Google should be able to push whatever it wants, on account of Chrome being open source. In other words, monopolistic behavior is fine as long as it's open source. I don't agree with that.

Also "Works best in Chrome" is hardly a self-inflicted wound. With the might of Google, that's actually a recipe for browser monopoly. This is the reason why Google pushed ahead with Dart despite the opposition.

How would your opinion change if Dart was just a cross-compiler to JS like GWT or CoffeeScript?

And if you have no problem with that, then why is there a problem with a Dart VM in Chrome? All it means is that potentially, Chrome apps will startup faster or run a little faster.

However, this is not really much difference than V8 vs other JS VMs. There was a time when V8 was way way faster than other JS VMs, so the same speed differential existed anyway.

There's a difference between "doesn't run" and "runs, but faster"

> If Google introduces something like Dart or NaCl, is willing to support standardization of it and even release a free implementation, but other browser vendors don't want to integrate it because they don't like it, that's their problem, not Google's. If that means popular apps become "Works best in Chrome," that is a self-inflicted wound.

I completely disagree.

First, if a technology is really bad for the web, are you saying other browsers should still implement it, just to not suffer the situation where popular apps work best in Chrome? That's bad enough as it is, but those popular apps are likely going to be Google properties - google.com, gmail. That means Google can basically force other browsers to adopt technologies as it sees fit, even if they are bad for the web.

Second, the WebKit situation on mobile exactly shows why this is wrong. If one popular browser adds special features and leaves them there indefinitely, in release versions of the browser - as WebKit does with its special CSS properties, and like Chrome does with NaCl and perhaps soon Dart - then people will develop for those features. If that browser then becomes dominant, those features will become a de-facto standard. Given Chrome's incredibly rapid rise, we should not assume it will not become dominant on the desktop, so this should concern us all.

Yes they may be forced to develop these things and to include them (if by being forced to we mean they would otherwise not having enough users) but I don't see that the browsers should have any say in what is good or not for the web. Browsers should focus on making the best browsers possible.

I am a web developer and I very much want it. I don't want Javascript to be the only game in town for front end programming.

But this is beside the point: Eich never asked if web developers want it or not.

"So many web standards advocates will cry foul the moment any vendor tries something out that has not already been standardized."

While some misinformed people might make this argument, I don't think you'll find anyone making it at Mozilla or other groups that actually participate in web standards development. We all know that the normal process is to implement first and then standardize. That's what Mozilla has done in the past with features like Geolocation and WebGL, and it's what we're doing today with new work like Audio Data, Game Pad, and Telephony. In fact, the W3C process requires two interoperable implementations before a specification can reach the Recommendation stage.

But there's a difference between a standard that is developed in the open, with the opportunity for all parties to provide feedback and influence development, and one that is developed and implemented in secret, and then deployed to production browsers and evangelized to content authors before anyone else has a chance to comment on it -- or even on whether the fundamental approach is a good idea. (And obviously these are two extreme ends of a spectrum, with a lot of gray area between them.)

The problem with many of Apple's "contributions" (and some of Google's) is that they never write specs for them or bring them to standards bodies. (This is the case for Aplle's -webkit-text-size-adjust.) In some cases they actively fight efforts to standardize them. (This is the case with Apple's touch events.) Or sometimes they do push specs to standards bodies, but only after a highly complex specification has been fully developed behind closed doors. (Google's Dart, NaCL, and Pepper have this problem to some extent.)

That's fine if what they really want is a proprietary ecosystem that works in their own browsers. But if they actually want these efforts to be part of the web, and if they want to be seen as stewards of an open web, then just implementing something is not enough. The web has lots of stakeholders, and the initial implementer needs to find a way to bring them to the table.

mbrubeck and I, who are good friends IRL, just had a spirited discussion about this over IM; I'll just repeat a few of my points here for other readers. :)

NaCl was first announced in 2008, along with a research paper and and source code (http://googlecode.blogspot.com/2008/12/native-client-technol...). The next year Google sponsored the NativeClient Security Contest that awarded prizes to security researchers who could break out of the sandbox. A lot of the design surrounding PNaCl has happened in the open (see http://www.chromium.org/nativeclient/pnacl). I don't think it's accurate to say that NaCl was developed behind closed doors.

Also, other browser vendors seem to oppose NaCl in principle, and it seems unlikely that any other process Google could have taken to get buy-in would have resulted in other browsers accepting it.

I agree that there is a continuum of behavior here; in the NaCl case in particular though, it really seems like Google has done the right thing.

"Never write specs"? I have to say, that some of the rhetoric from Mozilla seems a bit over the top. Mozilla doesn't seem to have had a problem dumping all kinds of Javascript and CSS features into Firefox over the years not all of it making it into draft spec proposals.

And how could NaCL have been done any better? It's been publicly published for years prior to the first shipping implementation. If Mozilla was interested, they could have put in requests for how they'd like to see NaCl and Pepper changed to better fit a Firefox implementation, or even a counter-proposal for running untrusted native code. But I've seen no such proposals, instead, just a blanket refusal to even consider it. I think there's a philosophical difference such that Mozilla isn't really interested in anything that isn't JS based, so suggesting that bringing NaCL to the working groups would have changed anything is somewhat disingenuous. It's kind of like hoping that Republicans would contribute to a National Single Payer healthcare plan design if it had been brought to the right committee.

It's a shame really, because proprietary native mobile platforms are hurting the web, and consumers don't really care, they just want fast apps which don't kill their battery or phone performance. Something like NaCL for mobile could preserve the DOM, CSS, browser security model, and other good things about the web, but replace JS with something much more performant for games, and this would go a long way to hedging risk against the proprietary native platforms. I'll take portal NaCl C-code calling into browser APIs and WebGL any day over Obj-C code calling into Cocoa that can only be installed through an app-store.

If the Web is to be around for decades more, the religion over programming languages needs to give way to polyglotism.

As on Dart, the language was released at version 0.1. The grammar isn't all that complex, the library is minimalist. It came with a early spec and source. It is not a "fully developed" specification. No one designs language grammars by committee. Languages are highly personal and opinionated. Usually one or two people design the overall structure of the language privately, and then release an initial spec, and it evolves from there.

When Mozilla developed Rust, the first time most of the web heard of it was Graydon dumping a gigantic codebomb into GitHub and it being covered on Lambda-the-Ultimate.

Which I think is perfectly find, but let's try and be a little level headed and fair, and not arrogantly think that what Mozilla does is uniquely better than what others are doing.

(I speak only for myself and I do not work on Dart, NaCl, nor Chrome)

It is unfortunate that you have intermixed good points with unfair and untrue attacks against mbrubeck ("disingenuous", "arrogant"); I know Matt in real life and he is none of these things.

I didn't accuse Matt being disingenuous, I accused Mozilla as a whole, primarily based on some of the comments of Brendan Eich when Dart was first released, because of somewhat contradictory statements.

Err, on second reading, I can see how it reads like I'm personally accusing him, my apologies, it was not my intent.

Not a problem, I am not easily offended. :) I was careless in my own comments and I apologize for being combative.

"Mozilla doesn't seem to have had a problem dumping all kinds of Javascript and CSS features into Firefox over the years not all of it making it into draft spec proposals."

Certainly, every vendor has been guilty here. I at least think that every extension Mozilla is currently encouraging use of is hoped to be a potential future standard. (And no, I'm not claiming this is unique -- some other vendors like Google are also good citizens in this regard.)

I also agree that proprietary platforms are hurting the web, and that the web needs new abilities in order to fight the threat. However, to preserve the openness of the web, new features need to be things that more than one vendor has the resources and motivation to implement; even if Mozilla and Opera thought Dart or Pepper were great ideas, adopting them would put is on a treadmill where Google (with its much greater resources) controls the speed.

And no, I am not claiming that Apple/Mozilla/Opera would have adopted Dart or NaCL if they were brought before the right committee at the right time. Part of getting early feedback means finding out if the community even thinks that your solution is something that's desired in the first place.

"When Mozilla developed Rust, the first time most of the web heard of it was Graydon dumping a gigantic codebomb into GitHub and it being covered on Lambda-the-Ultimate."

This is actually an interesting contrast between Mozilla and Google. The initial Rust code-dump was a barely-running proof-of-concept; it was shared with the world on the same day it was first shared within the Mozilla community. Zero lines of the current self-hosted compiler were written at the time. Most of the planned features (like the task system) were unimplemented. The first actual release (Rust 0.1 alpha) came a year and a half later, and that first release already reflected significant community feedback and contributions. For example, the packaging system in Rust 0.1 was written by a volunteer contributor who happens to be a Google employee. The stage of maturity at which Rust became open source was far earlier than Go.

Some things really are different at a non-profit like Mozilla, and the goal of opening everything up as early as possible is one thing that's unique among tech companies I've worked for or with. We don't always do it perfectly, and we do struggle with the need for people to do some types of design and experimentation without the world looking over their shoulders, but it's very much a unique core principle that affects all of our work.

SPDY is an example of a great idea, good enough that Mozilla is adopting it. Apparently, you're saying that, like SPDY, even if everyone agreed that Dart and NaCl were great ideas and great for the web, Mozilla still vote against them for political or aesthetic reasons. So part of "getting early feedback" as you alluded to above, is getting early politically motivated feedback. How is this a good thing? You should vote on technical proposals based on the intrinsic merit.

Almost all of the additions to the browser API that Google has made are desired to make it into spec proposals. There is no desire for Chrome to be the only implementor of Dart or NaCL.

I'm a little concerned what you're saying here. That if an idea is good, it won't be adopted anyway, just because someone else has greater resources to implement it. It sounds like Not-invented-here syndrome to me. No matter what technical specs are proposed, Google is always going to have greater resources to develop them.

Everything you say about Mozilla, I can say about my experience at Google. The first release of Dart was very incomplete, it didn't implement a lot of the type system, no missing method, some stuff like default method parameters didn't exist. It generally really slow and bloated code. It was in a barely working state. The DartC compiler which was implemented in Java was not a production compiler design to be used by people, it was a prototype for research. This idea that Google sprang a fully-formed already finished product on everyone is a gross exaggeration. The Dart VM isn't even in Chromium yet.

The self-hosted compiler (Frog, written in Dart) has been developed completely in the open.

Google is a large company with hundreds and hundreds of projects, so not all of them operate the same way, but many of them are open sourced very early in the process, and Googlers push to release source early for client side web technologies. Google has spent hundreds of millions of dollars on acquisitions which it immediately turned around and open sourced, like WebM, and the first Google Web Toolkit, and Instantiation's products (donated to Eclipse Foundation)

Really, I think the comments on NaCl and Dart are patently unfair. They were released practically as early as they could have been, NaCL took, what, 2-3 years in the open of development before it was finally pushed to production.

Ultimately, if standards committees get gridlocked over political and emotional stuff, instead over the merits of technology, you will see people start to ignore them. A similar thing happened in with OpenGL ARB where vendors who had less resources voted down and blocked companies like NVidia and ATI who were progressing chips at a faster pace. Ultimately, it took DirectX surpassing OpenGL to push it to finally start evolving faster again.

Yes, SPDY is a great idea, and of course I don't think that it -- or other ideas -- should be rejected just because of who proposes them or when. (As a concrete example, I am the lead editor of the W3C Touch Events spec and worked on implementing it in Gecko; it is an API that I absolutely believe should be standardized even though I strongly dislike how Apple originally designed and deployed it.)

I do think we agree on fundamentals though we may disagree on some cost-benefit analysis. All I meant about the "treadmill" is that some proposals have a much greater cost to implementers than others, and different organizations will weigh those costs differently. While some debaters may throw out "political or emotional" responses, I believe the decision-making process should be (and is, in the cases I've participated in) about allocation of resources, complexity of implementation, and the desired architecture of the web.

Technical issues and process issues are separate, but affect each other. When I say the process behind something like Touch Events was bad, I'm not giving that as a reason to reject the feature -- I think the feature should be accepted or rejected on its merits, like you said -- but it does mean that any technical objections that do appear will be much harder to resolve than they might have been. And as you say, Dart and NaCL have had better process behind them than Touch Events or various other extensions I could name. Dart and NaCL may succeed or they may fail (just as many of Mozilla's current experiments may suceed or fail), but at least their fates will be influenced by feedback from all stakeholders.

You are obviously better informed than I am about Dart and NaCL implementation. (Also, I accidentally conflated Go and Dart in my previous comment, which is embarrassing!) I apologize for any uninformed comments and defer to you on all the facts there.

I really don't want people to see this as a "Mozilla versus Google" thread. I hope everyone cares about standards issues because they help or hurt the web, not because they helps their favorite vendor or hurt a competing one. In the big picture, Google and Mozilla are close allies when it comes to web standards. As you say, the real risks to the open web are elsewhere.

I'd go further than that. I think the era of the technology standard as a formally specified, committee-managed, cross-platform entity is coming to an end. What we're seeing instead is the emergence of open-source implementations that are themselves the de-facto standard. This is particularly true of languages. Look how Apple has managed to run circles around the C++ committee in their single stewardship of Obj-C. Or think how ridiculous it sounds to think of a "standard" definition of Python or Ruby.

If it is true that we have moved past the era of formal standards, we will have lost something that is important. More important than most people understand.

If a technology has only implementation, even an open source implementation, that implementation will have all sorts of hidden assumptions and biases. No one will know about or understand those assumptions and biases until the time comes to use that technology in a new context.

In my work, I've dealt with multiple implementations of things like simulators, models and even languages. The additional implementations usually had some specific reason behind them (performance, alternative features, supporting different execution environments, different mathematical properties and so on). Nevertheless, the value of each alternative implementation (in terms of surfacing hidden assumptions, finding bugs, illuminating more general and flexible alternatives, etc.) went well beyond the specific reason that implementation was made.

These considerations are even more important in the case of software that is part of a large, chaotic, heterogeneous system like the Web. Imagine where the mobile web would be today if there hadn't been a W3C and, at the turn of the century, the Web really had been "what Internet Explorer renders" with no attempt to document and codify anything else.

Or, to take your example of Objective C, how much does Objective C matter outside of Apple's ecosystem?

Contrast this with:

- How much does JavaScript matter outside of Netscape's ecosystem?

- How much does Java matters outside of Sun's?

- Or, to take the granddaddy of examples, how much does C matter off of the PDP-11 and outside of Bell Labs?

None of those languages would be anywhere near where they are today without standards. On the other side of the coin, there is a reason languages like Python, Ruby and Haskell have had (and, each to a varying extent, continue to have) a "bus problem". Standards matter and even just the standardization process itself matters. I wish more people understood that.

How much does Objective C matter outside of Apple's ecosystem?

It doesn't need to. Apple has demonstrated the agility of a wholly owned, vertical stack. The pace of change has qualitatively changed since the days of language standards set by plodding standards bodies.

It depends on what your ambitions are. I'd agree that it doesn't matter much to Apple that Objective C isn't standardized (and not standardizing may be an intentional part of Apple's business strategy by increasing porting and switching costs).

But that also means that, no matter what its true potential is, Objective C will never be more than what Apple makes of it. It won't be seriously used outside of application areas Apple is interested in. It won't develop new features Apple doesn't care about. And if Apple decides to go in a different direction and/or runs into serious trouble, well that's it.

Apple has changed the world, but, unlike some other, standardized languages including the ones above, Objective C never will. If we lose the full potential of future great languages because they are limited by a single company and/or a single implementation, we will all be poorer because of that.

I used to think so but I've come around to the idea that the language is just part of the platform. Learning the Apis and idioms is always more work than learning a new syntax anyway.

I guess this is where we're actually parting ways. I agree that learning a platform, its APIs and idioms is important, non-trivial work, but...

As you might guess from my username, I'm a fan of more than a few languages where (at least I believe) the differences go well beyond syntax. That means I want standards because I want to know what I can (or could) take with me to another platform and what I can't.

The notion of a standard python is very real and meaningful, given that Jython, IronPython and PyPy all exist and have their differences from CPython. The only way you can write code that runs on all of them is with a standard.

> Or think how ridiculous it sounds to think of a "standard" definition of Python or Ruby.

There are ongoing efforts to standardize a quite sizable subset of Ruby in Japan for later ratifications in ISO. Ruby is also de facto standardized by the test-suite RubySpec.

It is not ridiculous at all, in fact reference documents such as RubySpec have been one of the main reason for which you can move from a Ruby implementation to another without many issues and without having many compatibility hacks included in each compiler/interpreter.

Or think how ridiculous it sounds to think of a "standard" definition of Python or Ruby.

Really? It doesn't sound all that ridiculous if you know cpython is the reference implementation of the language python, which is extended by the PEP process.

My point exactly. There's a reference implementation, not a thousand page standard document.

>not a thousand page standard document

It has those too. And IIRC the w3c requires a browser to implement a reference version before something is standardized there.

There are plenty of W3C standards that WebKit is not willing to support. That's fine. Some of the standards just suck.

There are also plenty that it purposefully implements incorrectly (CSS Selectors, for example). That's a lot less fine.

> There was a time when the "Internet" could if you squinted right be defined as "the collection of people who make fun of that process and do the opposite"

Are those the same guys who brought FTP upon us? BGP? 32-bit IP addresses?

There were few moments where it would be better if somebody said, wait a minute, we need to think first, then code.

Yes, and this is the same shotgun process that managed to avoid congestion collapse. Not by spending 6 years fretting about it, but by writing the TCP stack in 4.3BSD-Tahoe.

32 bit addresses were a good idea. It's easy to look back on it now and say "they should have used 128 bit integers", but until recently, 32 bit addresses were the largest addresses that worked as scalars in all mainstream platforms.

>FTP upon us? BGP? 32-bit IP addresses?

Those are effectively entrenched de facto standards, but its not the fault of the creators, its the fault of the people unwilling to move with the times. I'm not cool enough to run an AS, but I know for the other two better alternatives are already here. As Vint Cerf has said, 32 bits was enough for an experiment, it just never ended.

IPv6 and SFTP are here and _in use_ by anyone who isn't clinging on to old tech like people clung to IE6.

>we need to think first, then code

Thinking first, implementing and then improving is how technology advances. Iteration is mandatory, otherwise you're stuck in a waterfall-type scenario. Without it, I sure hope you enjoy riding your horse down to the local Gutenburg press to pick up this week's HackerNews and read my comment there.

> its the fault of the people unwilling to move with the times.

It's the network effect. Once something is "out there", it's really, really hard to get rid of it. If the first version that reaches critical mass has unfixable flaws, you'll be stuck with them forever.

I don't blame Vint, he couldn't have idea how successful the experiment is going to be, but I can't read HN over IPv6.

But now we know how web standards spread and stick around forever.

HTML5 special-cases parsing of <XMP> element. CSS has quirks mode, almost standards mode and standards mode and two syntaxes for clip().

If you read the article referenced, web-kit isn't being picked on especially. The concept of using vendor prefixes is being questioned.

There's a valid rationale to not using vendor-prefixes, because it could potentially add confusion and erode standards.

From my understanding, in effort to display web-pages at their best, some browser manufacturers are starting to honour the browser prefixes used by other vendors. I can see how this could create problems.

It's a little hard for me to avoid the impression that W3C partisans hate vendor extensions because they make it easier for websites to handle cross-platform support for extensions. -moz-border-radius and -webkit-border-radius is ugly but nothing more; conflicting definitions of "border-radius" is a showstopper.

Which, if you're cynical, is how the W3C probably wants it.

The alternative isn't to stop using the properties - it's to stop using the vendor prefixes. That's all that's being suggested.

I think your aggressive anti-W3C stance is really misguided. They're not the bad guys.

If you actually read the article referenced in the dcurt.is article, you'd see how vendor extensions are actually becoming something more than just ugly. Which is the whole point of the suggestion.

The argument in that article is adequately summed up by Dustin Curtis. The problem with vendor extensions is that when one browser ends up dominant, the others eventually end up having to implement that browser's extensions.

What I don't get is how this makes that browser a "monopoly". The only thing Webkit won was the string "-webkit" in the CSS property. It's the monopoly that gives the monopoly to Webkit. Not CSS vendor prefixes. Webkit has a virtual monopoly among mobile browsers.

WebKit has an advantage with WebKit prefixes when the prefixes in question have no spec. As DarkShikari said when criticizing the VP8 spec, code is not a spec.

Let's see if we can't tell the difference between -webkit-transform-style, a fiddly CSS property that to be useful has to be documented enough for web developers to employ it, and the interpolation filter embedded in the guts of a black box video codec.

The problem isn't that one browser is dominant. The problem is when Apple develops a feature entirely in secret and then doesn't bring it to a standards body, forcing other browser vendors to reverse-engineer that feature. Apple is shipping those features, and evangelizing them to developers, before the other browser vendors can even get started reverse-engineering them (which they shouldn't even have to do in the first place).

Now, you could argue that that's in Apple's interest, and you could argue that "it's their browser and they can do what they want with it". But you can't dispute the fact that they're not helping web standardization in the mobile space. Which is a problem for anyone who wants a more standardized web.

Good point. Webkit even changed the way it handled gradients about a year ago to be in line with standards (at the time I liked the webkit way).

Reply on news.ycombinator.com