Hacker News Reader

Even if it were legal (which it should be IMO), I can't see anything other than a cat and mouse game happening where the primaries change their authentication patterns on a daily, hourly or minute to minute basis. Consider all the things that facebook does to mess with scrapers right now - every single letter in a div, multiples that are hidden, etc. Its a complete mess.

"Forced open APIs" is an interesting idea in terms of actually bringing value to users - its almost on the level of nationalizing a company though in terms of being anti-business. Why Facebook would front the cost of running a service they can't monetize though is a big question.

Forcing open APIs isn't anti-business. It's pro-competition and pro-startup. The only thing it's anti- is big incumbents, which is the point.

We've gone from social media being a relatively unmonetized public good to a heavily monetized semi-private good. I hope we can find some middle ground eventually - though the "problem" with unmonetized services is they go away because nobody can afford to run them (Wikipedia being a good example of one that is treating the line as a non-profit).

Isn't fixing that the promise of distributed apps?

> every single letter in a div

That's... I... Well. :0

That alone is enough to assure me my conviction to never support them beyond a token account was justified.

...though I may formally delete my account now. That's just... I'm going to be sick. That's Hostile Architecture if I've ever seen it.

https://en.m.wikipedia.org/wiki/Hostile_architecture

I'm growing increasingly weary of this type of offensive programming. I've never felt or understood the need to actively make something as hard to consume and process as possible. Again though, it really speaks to Facebook's view of their users: 'Look, but don't touch!'

Then again, I tend to build by the maxim 'it is vain to do with more what can be done with less'.

How dare they. They should put every letter in a <span> instead, since letters should be inline. Heathens.

Do they really do this? I would totally believe it but I don't see it...

Next time you are on facebook, just right click and inspect element or view source code.

Easy to get around. Just use "innerText" and it'll generate whatever text is inside. They do have plenty of other tactics, though.

They also will write the divs out of order, put in "fake" letter divs that are hidden or on a different z axis, all sorts of fun stuff.

I think your focus on the CFAA is somewhat missing the point. The assertion being made is that Facebook and Google (and perhaps one or two others, like Amazon and Twitter, maybe) have become so big and so powerful that they are able to act like the trusts of yore, leveraging their data to stop competitors from having any realistic chance of disruption and even affecting other peripherally-related industries. If we suppose open data access to be a remedy, in whole or in part, then "don't allow companies to use the CFAA to stop third-party data scrapers" is an insufficient proposal.

Suppose the CFAA wasn't applied so indiscriminately, but Facebook still didn't want third-party clients to access their data and there was no requirement for them to do so. So they make sure there's no publicly-available API, at least with any terms of service that let you build a true third-party client. Assuming there is still some legal mechanism to make unauthorized access to computer resources illegal, then tapping into Facebook's private APIs counts. Doing what 3Taps and PadMapper did -- just scraping data -- can be a Terms of Service violation (as can the private API access, presumably) and could conceivably even be charged as a copyright violation. It's asking for trouble. It's also potentially a constant cat-and-mouse game. And, last but not least, there's no guarantee that data scraping will actually get you the data that you need to do a good job of building a new business on top of Facebook's social graph.

What would be better, if you accept the original premise, is not new regulation, per se, but antitrust action specific to these companies forcing them to build open APIs. Simply saying "hey, Facebook can't sue you for data scraping anymore" just won't cut it.

I'm not claiming it's sufficient. Perhaps it's not.

But it seems mistaken to use anti-trust regulations to break up or regulate companies when there are still laws on the books that help facilitate market dominance. Repeal those first and see what happens before doing anything else.

Any legal rule that makes legitimate HTTP requests illegal (i.e. requests whose purpose is not solely to commit harm ala a DDoS) ought to be repealed.

> Rather than trust busting, why not start with relaxing the federal laws which forbid the development of third-party applications?

Cute phrasing, but what this really means is "why not force anyone who develops a product to allow others to capitalize on it?"

> It seems premature to pursue trust busting action while we still have federal laws that encourage and cement the dominance of a single provider.

You mean "allow a product to exist as a unified product". "Provider" is 100% misleading language.

>Cute phrasing, but what this really means is "why not force anyone who develops a product to allow others to capitalize on it?"

. . . why not though? I don't see how permitting monopoly rents in perpetuity is an economically or socially desirable outcome.

Facebook is a product, not a monopoly.

I said "monopoly rent." It's a term with a specific definition that you would do well to understand.

Pretty much any definition of "monopoly rent" I've found does actually require the existence of a monopoly. Also, there's no reason to talk down to people.

https://en.m.wikipedia.org/wiki/Economic_rent#Monopoly_rent

>I've found does actually require the existence of a monopoly.

Monopoly is defined loosely and based on the idea of market power, not being a sole provider. The Wikipedia article you cite even mentions Microsoft and Intel as examples, and they had less market power over their fields than Facebook does.

Monopoly rent implies (near) monopolies, which Facebook is not.

What's more, this is a conversation around individual users. Facebook's economic rent is paid by their customers, which are advertisers and not the subject of this conversation.

>Monopoly rent implies (near) monopolies, which Facebook is not.

Except it is, because it's able to leverage its market power to dictate terms to the market. If you prefer the term "oligopoly" rent you can do that, but there isn't much point. Their network effect and capacity for lock-in allows them to collect economic rents, which is the point here.

>Facebook's economic rent is paid by their customers, which are advertisers

No. Loss of privacy, surrendering your data and information, and platform constraints are all forms of rent paid to Facebook by consumers.

I think you're misunderstanding. I'm not proposing that Facebook be _forced_ to allow third-party access, which would carry the risks you've pointed out. I'm simply proposing the federal government not aid them in their market dominance.

Facebook could still prevent third-party applications from functioning, but it would have to do so by methods inside of the HTTP specification rather than legal ones. For example, it could simply ban users who authenticate via a third-party app.

It seems like the argument is for large-scale IP protection reform. What one could do there is create an actual new class of asset in "data" with different rules so that traditional asset law doesn't apply.

What's the difference? If it's desirable to reserve the right to disallow third-party frontends, why push the burden onto them (and everyone else) enforcing it through technical measures?

Because with technical measures, the cost is allocated to the party who benefits from vendor lock-in, rather than to the public.

>Cute phrasing, but what this really means is "why not force anyone who develops a product to allow others to capitalize on it?"

Or maybe that could be the form of trust-busting: by default, you don't have to give the access to the data, but if you're found to be an abusive monopolist, then you can be forced to.

> It could, for example, finally be possible to see a time-ordered view of your friends' posts (Facebook doesn't allow this since it reduces engagement).

I view this the same way as I view ad-blocking.

Building on top of Facebook only strengthens their monopoly and does nothing to reward their competition.

And given that Facebook can break such "apps" on a whim, via simple and justifiable changes to their UI or APIs, you haven't gained anything, while reinforcing their monopoly.

Facebook allows users to easily export their data as a compressed archive.

Why not have the third party app access exported archives and not Facebook servers?

This way users could at least be made aware of the features they are missing. Ones that might be better for users but not necessarily better for Facebook.

For example, organising and searching through past data could be made much more efficient, reducing the time users need to spend on the Facebook website/app.

> Facebook allows users to easily export their data as a compressed archive.

> Why not have the third party app access exported archives and not Facebook servers?

For the most part, those archives only contain data that the user has personally uploaded. They contain little to no data about their friends or their friends activity. A third party app that only works with a user's archive will not be able to function as an adequate Facebook alternative.

I was thinking that users would share relevant portions of these archives with their friends, etc. directly, without a middleman.

They have already shared the content of the archive via Facebook.

Nor was I thinking of a "Facebook alternative" (neither was the parent comment). What I was thinking is that one could demonstrate certain features in an app that one does not find one Facebook.

Users would then become familiar with features that are possible but that they are not getting on Facebook.

Or, why not build open source alternatives that are as good as Facebook?

Back in the day AOL was the Facebook. The first browsers were ehh, but the idea of owning your own domain really appealed to people and the Web killed AOL which is no longer around as the world’s centralized social network.

That’s what we are doing at https://qbix.com . Watch the video there. We are inspired by Wordpress which powers 30% of all websites - we just wanted to build a true alternative to Facebook, not just a blogging platform.

Example app: https://davosapp.com

PS: I see a few steady downvotes every time I suggest that we need an open source alternative to Facebook and link to meaningful efforts in that direction. I would definitely prefer to hear what is wrong with what I said, rather than silently downvoting.

> PS: I see a few steady downvotes every time I suggest that we need an open source alternative to Facebook and link to meaningful efforts in that direction. I would definitely prefer to hear what is wrong with what I said, rather than silently downvoting.

I don't think you were downvoted because you suggested an open source alternative to Facebook, but rather because the bulk of your comment was a plug for your own project.

You'd probably get a better reception if you made your point, and ended your comment with a very short plug along the lines of "PS: I'm working on one at https://qbix.com."

How would you deal with the risk of Cambridge Analytica style entities using all the third party app ability?

A third party app can access FB servers directly and scrape content as needed. No law prevents this. You'd just be using a custom user agent.

I agree with this - in the recent discussion on Slate Star Codex about Thiel's views, the point was made that outsize success and profits are generally a sign of lack of competition.

To me, after a reasonable period for them to capitalise on their innovation/higher quality execution, any company making huge profits in an area with few competitors should be subject to a range of measures designed to reduce barriers to entry to make it easier for new competitors to emerge.

Some natural ones would include rules about standards, providing apis, user portability, etc.

Indeed, we wouldn’t want to incentivise people to produce wealth and improve people’s lives. That way inequality lies.

No actually. We do want to incentivise that behaviour, which is why I suggested a grace period for extracting the outsized profits.

What we dont want to incentivise is regulatory capture, entrenched power, rent seeking behaviour.