Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

No.

They could matter, if they built a browser with competing market share. It's not a static equilibrium.



Did you look at the list of members? Who on the list is going to be building their own browser and what would the business case be? Out of the big four, only two of them even thought it made sense to build a browser from scratch. Apple and Google started off with KHTML.

EDIT:

To clarify. Who is going to be building their own rendering engine instead of taking an existing one - 3 of the 4 are open source - and building a browser on top?


The lineage of the rendering engine doesn't matter. What gives Apple and Google and Microsoft command of the standard is the fact that they mediate access to web pages; they own th actual customers. That's what matters.


Firefox doesn't own the customer.

But creating a rendering engine from scratch is hard and there is no business case for anyone doing it from scratch. Apple didn't (they tried with CyberDog ages ago) they used KHTML to create WebKit. Google didn't either, they started with WebKit. Opera gave up on their own rendering engine years ago.


Why is "creating a rendering engine from scratch" the bar here? A new player could fork Blink or WebKit.


If you fork an existing rendering engine, you're rather implicitly using a WHATWG DOM. In order to use W3C DOM 4.1, you'd either have to modify a WHATWG DOM renderer into a W3C DOM 4.1 renderer -- which is a bit like starting with the emacs source to build a vim clone -- or to build your own. The inertia of forking an existing project pushes you towards the WHATWG implementation, not W3C. That's the point of forking. You get to preserve the forking.

Further, the W3C's argument historically has been to make the DOM easier to implement from the ground up (XHTML strict) compared to the overall rats nest of HTML5, which, as far as I'm aware, still is not fully supported anywhere [1] and is very loose about document structure errors etc. So, if you're planning to implement the W3C's DOM, it makes sense that you're agreeing at least somewhat with the W3C's historical philosophy about what the web should look like and how it should behave, so you're more likely to be concerned about the implementation difficulty of HTML5.

1: https://html5test.com/results/desktop.html


I think you have the cart before the horse here. WHATWG is just a codification of what is. If we didn't have WHATWG, you wouldn't be freed from the burden of supporting current webpages, you just wouldn't know what that burden entails. At any rate, refactoring Blink to meet W3C's spec has to be easier to write a renderer from scratch, even if you don't care about writing noncompliant webpages, or as they're usually known, webpages.


If that's the case:

1a) Why would it matter that Microsoft, Google, Mozilla, and/or Apple object to W3C DOM 4.1 if they don't implement it?

1b) Why would Microsoft, Google, Mozilla, and/or Apple care enough to object to W3C DOM 4.1 if they aren't implementing it? Why would they even give any effort to a competing specification and just allow it to die from inactivity?

2) Why does what is in W3C DOM 4.1 matter if the high 90s percentage of users are served by a browser in the WHATWG DOM camp? This could probably be condensed down to "Why do W3C's specifications matter at all" really.


> Why would it matter that Microsoft, Google, Mozilla, and/or Apple object to W3C DOM 4.1 if they don't implement it?

It matters to the utility of the W3C DOM spec that it doesn't represent either what browsers have implemented or what they will implement.

> Why would Microsoft, Google, Mozilla, and/or Apple care enough to object to W3C DOM 4.1 if they aren't implementing it?

They care enough because they want the W3C, if it is going to write purported web standards, to do something that won't confuse developers and lead to browser vendors fielding complaints from developers who mistake useless W3C documents for something meaningful.

> Why would they even give any effort to a competing specification and just allow it to die from inactivity?

They don't want to have a competing specification, though they do not seem opposed to having a specification with a different focus but consistent with WHATWG to the degree dictate by the purpose.)

> Why does what is in W3C DOM 4.1 matter if the high 90s percentage of users are served by a browser in the WHATWG DOM camp?

The idea is not to have opposing camps, though if W3C insists on making it an opposing camps situation, thst becomes a real issue.


Exactly. They don't want any competing products so preventing the development of a standard they don't control is obvious.


If this is about preventing competition in the browser space, then why doesn't one of the browser makers with lower market share defect from Google's position?


No, you start with a WHATWG DOM. What you do after that is up to you.


Creating a browser isn't the only part of mediating access to web pages. In different senses, Digicert, Comcast, Akamai, and Cisco do that as well.


Fair point, but I think it is somewhat orthogonal to the discussion. I don't imagine Cisco cares which DOM spec is used in rendering the application layer bytes of the packets it routes.


The browser is the only part of the web-access stack where the user has any choice. I suppose they can choose the ISP as well but that effectively only changes access speed, possibly.


Nonsense. Web standards are supposed to be de jure, not de facto.

Once upon a time Microsoft had 90% of the browser’s market. We created web standards in order to prevent monopolies, such as the former IExplorer, from holding the market hostage. That’s the whole reason behind web standards.

And yes, they matter even with an IExplorer that has 90% market share, because governments can and do enforce adherence. That’s also the reason for why Microsoft came up with OOXML, ODF being a threat even with a tiny market share.


It doesn't matter what is suppose to happen. In reality. If none of the popular browsers support the standard - it doesn't matter.

Governments are not going to force every major browser manufacturer to support a standard.

That's why W3C lost relevance.


Web standards are supposed to be de jure, not de facto.

HTML5, in large part, was created to do exactly the opposite -- formally set down in writing all the de facto quirks of HTML as actually used, parsed and rendered in the real world, instead of continuing to prescribe behaviors which didn't match observed reality.


You and I lived a different history then, because if what you're saying is true, then ActiveX should have been standardized.

We've got no ActiveX, so your claim is false. Mozilla actually could implement ActiveX. They refused to do so.

Also, lets not forget that IExplorer 6 had incompatibilities with the standard, including XMLHttpRequest, even though Microsoft invented it.


For the most part it was documenting the common subset of how the browsers actually worked. Only one browser implemented ActiveX or ever wanted to so it isn't in the spec.


Nothing in your comment actually refutes anything I said.


"Who is going to be building their own rendering engine instead of taking an existing one"

Just for the note: I did - https://sciter.com

It was not meant to render all possible pages from Wild World Web but it renders HTML5/CSS3 (some subsets but still).


Impressive! So what are your thoughts on the technical merits of W3C's DOM approach? You agree with Google/Apple or is this just a case of them using their power to lock down the market?


Awesome! How would you compare your product to Electron?



TL;DR: It's small and performant.

But the thread is an interesting read.


Holy Crap that is bloody impressive!.


> if they built a browser with competing market share.

As if building a competitive browser isn't hard enough. You then have to convince people to use it. Considering the walled gardens 3 of the big 4 are erecting around the platforms the control, that seems neigh impossible.


Users should have a voice too.


We definitely believe users should have a voice in the WHATWG, and thus in guiding what browsers implement. We strive to maintain an open and welcoming community; this has brought a lot of good ideas to the table.

A few years ago I gave a talk on this. https://www.youtube.com/watch?v=hneN6aW-d9w . I hope it's not too embarassingly outdated now :)

In particular, unlike the W3C, we do not require membership fees (https://www.w3.org/Consortium/fees?countryCode=US&quarter=04...) for participation.


Right, that's my point :) But thanks for making it clear for those reading.


They do: they pick which browser to run, and thus give power to.


That's not a free, expressive choice. One shouldn't expect that the spectrum of browser maker's choices align with user's preferences. Also they might use a browser because a website requires it, not because it aligns with their preferences, or because it was built into their phone, etc. etc.


Hmmm... Do you remember IE6 days? It was the best browser at that time - at least in respect of user base.

Yes, they did a lot of innovations there we all use now. Most notable - the whole AJAX idea was born there.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: