[groby_b]

It's kind of amusing to read an article complaining about the walled garden of Facebook, and at the same time demanding central ownership for web technology.

It's not only a possibility that a centralized web will be bad, it's inevitable. A central owner has _no_ incentive for improvement. The only reason e.g. Cocoa keeps improving is because it needs to compete.

[joehewitt]

Has Linux stopped improving? Has Python stopped improving? Has jQuery stopped improving? It all depends on who the owner is.

[groby_b]

Linux doesn't own the OS market - far from it. Python doesn't own the language sector. jQuery doesn't own the JS library sector. They are all independent projects, and it's _good_ for them to have a BDFL.

Asking for one for the entire web is not a good plan because it would exclude competition. That's where the flaw in the argument is.

And while it'd be nice if browser vendors could just go off and improve along (from a browser vendor's point of view), it absolutely sucks for every single developer and the users. Surely, you remember "this website runs only on...". This is exactly what you see in the apps market - applications only running on certain devices, locking in the users. Sure, for you as a developer it's convenient. For users, it's a bad idea.

[joehewitt]

My point wasn't that Linux or Python or jQuery own their sectors. I responding to your claim that a single owner promotes stagnation.

It's important to distinguish the Internet from the collection of web technologies. The Internet is the thing that needs to remain open and uncontrolled.

[groby_b]

My point is that a central owner _without competition_ stifles innovation. There's only one web, so a central owner would be a bad thing - for the web. To quote your article "To thrive, HTML and company need what those other platforms have: a single source repository and a good owner to drive it"

The only other way to read that is that you're advocating splintering web technology into HTML-Moz, HTML-IE, HTML-Cr and so on. While that might make for a better developer experience for somebody who develops apps for _one_ of those platforms, it's a disaster in terms of interoperability. Which is kind of one of the key points of "The Web".

We were heading down that path for a while with browser-specific extensions, and it lead to large problems. For users, because they needed to have "just the right browser". And for many backend services, because actually extracting data from web pages required understanding all competing standards.

I care mostly about the users. The web of yore, with "best viewed with" stickers was a debacle in terms of UX.

[ootachi]

Python pretty much has stopped improving; the world is still largely on Python 2.x. In fact, there was a highly-publicized moratorium on new features that lasted for a long time.