[stroboskop]

'Self-defense'? The post's key quote sums up Google's relation to RSS:

lead with a compelling user experience first and then build an API from there, an API which may be based on open standards, but only if it’s a means to an end.

Although an open standard, RSS is a special case: The DNA of RSS is incompatible with the data-greedy centralization enforced by the Google, Facebook, Twitter and their likes.

[Gormo]

Seems like a variant of the old MS strategy of "embrace, extend, exterminate". The open source movement was in part motivated by a desire to limit Microsoft's control over software users in the nineties.

Perhaps it's time for an organized "open internet" movement to build protocols and communities that resist the tendency toward centralization of data and control of user experience that's increasingly evident in the services offered by the big players (and even the smaller ones - today, I wanted to post a comment on a blog article, only to discover that the latest version of Disqus has disabled OpenID support).

I think walled-gardenism on the internet has far more dangerous implications than closed-source software ever did, and it's really sad to see this spreading meme of building walled gardens as the only path to commercial success infecting Google.

[mmahemoff]

As the OP, I agree it's sad that decentralized standards haven't "won" in the way they were expected to 10 years ago; it's the reason we now have Facebook, Twitter, and Google connect buttons instead of just Open ID.

The question is, when you say "it's time", how do you make it happen? Google tried to make it happen with standards like OpenSocial, but the trade-off for increased flexibility was often poorer UX, and meanwhile they watched developers jump onto standards that were more closed, but had many more users.

There's certainly a sweet spot where open standards meet a mainstream user base; the web and HTML5 overall continue to do fine, notwithstanding heavy competition from the more closed native platforms. This is very much due to the great amount of innovation amongst browsers and web apps, both of which touch the user directly, and less because users care about open for open's sake.

So my suggestion is if you want to encourage open standards, focus on the user first.

[Gormo]

> The question is, when you say "it's time", how do you make it happen

I'd say a good starting point would be something akin to the GNU project for protocols and services. Stallman's work made alignment with a particular set of principles the overriding goal of software development, and as much as this is often regarded as an extreme position for prioritizing ideals over the practical value of the software, it's hard to deny that it certainly shifted the 'center' of discourse to a point that gave enough weight to user freedom to enable a thriving ecosystem of open-source software that satisfies both practical use cases and the ideals of software freedom well enough.

We've seen a lot of one-off projects that have attempted to create distributed, user-centric services and protocols - OpenID, Diaspora, etc. - but these haven't aligned into an overarching "open internet movement" where projects build upon each other's work, or endeavored to promote a unified vision of the open internet.

Basically, the principles of an open internet ought to be articulated in a coherent statement of purpose - something akin to the FSF's "four freedoms" - and attached to some effective branding. The ideal needs to become a meme.

Marketing the idea of the open internet shouldn't be too hard: there are already plenty of examples of people's lives and workflows being severely disrupted by service shutdowns, business-model restructuring, security breaches, and so on, to which outsourced non-commodity software-as-a-service offerings are uniquely susceptible.

When we look at the kinds of practical concerns that have lead to this structural milieu - i.e. the situation in which service vendors are actually able to shoehorn their users into walled gardens - the single factor that pops out is the fact that the service vendor itself is in control of the platform on which the server operates, and can therefore modify the structure of the application or protocol without restraint.

The first order of business for a practical solution, and the first kind of product that ought to be developed, is something that breaks that combination, and gives users a level of control over the web applications they use that's akin to what they expect for desktop apps.

What if everyone had their own VPS, with a user-friendly UI to install and configure server-side applications, that give them the ubiquitous access and ease-of-use they're currently getting from webapps while still leaving them in control of their own user experience, and allowed them to choose what products to install, what versions of those products to use, and what features to enable?

A VPS-as-end-user-platform model would break the current platform/service combination that lends itself to walled gardens, and allow the VPS providers to compete on price and quality of their commodity service, while application developers would compete to encourage users to install commercial or free web-based RSS readers, OpenID implementations, social-networking nodes, email clients, etc. onto their own VPS instances.

[anon1385]

>As the OP, I agree it's sad that decentralized standards haven't "won" in the way they were expected to 10 years ago;

I don't know what most people expected, but it seems like there were at least some people complaining about the direction the web (or internet as they it called back then) was going even in 1997:

http://www.arachnoid.com/freezone/

The root of the problem is commercialisation, and a lot of people are to blame for that, including the creator of the site you are currently using.

[Gormo]

I don't think the problem is commercialization per se; it's the short-sightedness and narrowness of the current commercialization strategies that are the source of the problem.

Google became a multi-billion dollar company by supporting and contributing to the open internet over the course of ten years; their current worrying tactics are very recent. So we know that it's very possible to be wildly successful without undermining your customers' interests (and in the long term, undermining your customers' interests is almost always unsustainable).

The problem is that we've got big players like Google and Facebook who have become risk-averse as they grown, and, having maximized the potential of their original founding visions, have shifted into consolidating their positions in order to preserve the status quo at the expense of others. This is a pattern that seems to recur again and again in the industry.

The way to break it, of course, is to be the source of the creative destruction that undermines the status quo - few large, vested enterprises are willing to do this, though, which is why we see them ultimately becoming dinosaurs who are displaced by startups operating under new paradigms.

I'd hoped that Google, given its nature, would be the one organization that might be able to avert the pattern, but I guess not; they should be doing exactly the opposite of what they're doing now, and support a wide range of products and services, and looking for innovative monetization strategies for products that aren't immediately profitable. But instead, they're going for ultra-focus on what seems to work in the here and now, and trying to entrench the status quo, which will take them down the well-trod path to eventual failure.

[mgkimsal]

You'd actually need to build the software that gets used, not just protocols.

The majority of the browser usage is with three pieces of software: IE (Microsoft), Chrome (Google), Safari (Apple). None of these players really have a huge stake in 'openness'. We're computing at their mercy right now, and if/when they decide to adopt some new protocols (or drop support for others) we all just have to suck it up. Move to Firefox is a good option right now, but might not be in a few years.

[Locke1689]

No, it's more like IE, Chrome, Firefox.

[mgkimsal]

depending on what target you're looking at, yes, for now. firefox has nothing on mobile right now - it's chrome/safari on mobile by a longshot, and mobile is the hot growth area. But yeah, point taken.

[Gormo]

> The majority of the browser usage is with three pieces of software: IE (Microsoft), Chrome (Google), Safari (Apple). None of these players really have a huge stake in 'openness'.

Sure they do; all of the browsers support the same standards in HTML, CSS, JavaScript, etc. The few remaining points of contention between browsers, like HTML5 video formats, are relatively trivial and not significant with respect to the services that are starting to act like walled gardens.

The problems come from the applications and services that use the web itself as a platform, and attempt to "embrace, extend, and exterminate" open protocols, like RSS and OpenID, in order to lock users into relying on proprietary APIs instead of open standards.

As the previous commenter pointed out, the threats today come from Google, Facebook, and Twitter, not from the traditional desktop software vendors. (Only Chrome is really concerning here, since they're attempting to use Chrome as a way of shoehorning users into Google services, much in the same way that Microsoft leveraged their OS dominance in the '90s to boost their desktop applications, especially IE.)