Ok, so instead of breaking you up into three companies, we'll break you up into 4. The fourth can own the messaging infrastructure and license it out to anyone interested.

I can find a legal solution to your technical blockades. Be careful what you ask for.

>I can find a legal solution to your technical blockades. Be careful what you ask for.

legislative... senator.

I think he is referring to the Justice Department

This is fun to read but also implies that we have representatives competent enough to even understand things at this level. Maybe they don't have to understand it, I doubt they even consult someone who could help them understand this.

That's the role of staffers, though the rise of the lobbying industry has perverted this somewhat.

Some folks in congress are savvy about technology however, such as Wyden.

This isn’t new. Telephone infrastructure, power grid, roads and railways in some countries are using this exact pattern with some success.

I laughed at this because I thought the same thing.

Conspiratorial or no, that line of argument won't fly. Saying everything is so integrated that it can't possibly be decentralized is basically proof of monopoly.

It would also be a hilarious argument to make after previously convincing the European Commission in 2014 that there is no technical way to integrate Facebook Messenger and WhatsApp user accounts. [0]

[0] https://ec.europa.eu/commission/presscorner/detail/pl/IP_17_...

> convincing the European Commission in 2014 that there is no technical way to integrate Facebook Messenger and WhatsApp user accounts.

What was the argument there? It's not like that's solving the halting problem.

I think “technical” here just means something like: one uses phone numbers and the other uses emails. The current technical solution doesn’t make it straightforward to merge. Evolving the technical solution would require product and user facing changes. And probably a lot of edge cases

That sounds like a tech interview question

I am not sure why a shared messaging infrastructure implies merging accounts is easy/possible. It's fully possible to have messages pass on a common infrastructure between different entities that correspond to messenger/whatsapp accounts.

I'm not saying I necessarily believe facebook here, but I don't understand how they are contradicting themselves here.

> Saying everything is so integrated that it can't possibly be decentralized is basically proof of monopoly.

That doesn't make any sense to me. I would expect different applications from the same company to share the same underlying infrastructure. Whether or not they are a monopoly is orthogonal to the issue of integration.

> Whether or not they are a monopoly is orthogonal to the issue of integration.

So, I get you here and what you're saying. But companies at this scale argue they're "independently managed" or whatever.

With that context, for Facebook, there's really no way of saying that these 3 different brands are independent when it comes down to it.

It'd be much like Procter & Gamble saying their 100s of brands were all really different institutional entities.

At the far end of this reasoning three separate companies that all use AWS share the same underlying infrastructure and all go down when AWS does. It doesn't mean they're not independently managed.

And the same could be said for the electrical grid, the post services, roads, etc. I agree that infrastructure and monopoly are two very distict concepts.

But there aren't 3 different Amazon marketplaces--owned by Amazon--built on that infra. And AWS is available standalone.

Here we have core infrastructure that isn't sold to anyone else and powering 3 supposedly different social messaging services.

The key isn't the backend, it's the supposed independence of the frontends.

No, it’s relatively easy to go from one cloud platform to the other (thanks god for terraform). To go out of fb infra? I can’t even imagine.

> Saying everything is so integrated that it can't possibly be decentralized is basically proof of monopoly.

This doesn't matter from the standard of antitrust law. What most fail to grasp is that being a monopoly is not illegal. It's only illegal to use monopoly power to harm consumer welfare.

In this case integration between services is a pretty strong defense against this. Facebook can argue that consolidating the infrastructure allows them to deliver each product at a lower cost and/or higher quality than if it was served independently.

If the court accepted those facts, then Facebook would have a ironclad argument against being broken up under the consumer welfare standard.

> > Saying everything is so integrated that it can't possibly be decentralized is basically proof of monopoly.

> This doesn't matter from the standard of antitrust law. What most fail to grasp is that being a monopoly is not illegal. It's only illegal to use monopoly power to harm consumer welfare.

Citation needed? I thought, in the US at least, that the government will go after you if you use your monopoly powers to enter into other markets. For instance Google surfacing its own, other products as the top results in search.

It depends on how the Justice department feels, to some extent. GM in the 50s and 60s actually kept an eye on their market share, and sometimes would kill/cutback projects or products that may have threatened going over a certain percentage (around 60%, IIRC.) At that time merely going over that number would have been enough to start an investigation.

Some would argue though that back then DOJ 'cared' more.

If you're curious about how this changed, "Amazon's Antitrust Paradox" by Lina Khan article is a wonderful read: https://digitalcommons.law.yale.edu/cgi/viewcontent.cgi?arti...

The summary is that the Chicago school of Economics promoted the idea that in a competitive market, predatory pricing is not possible. Thus, when you see an industry structure emerge (e.g. duopoly), you should assume that it is an efficient competitive outcome. After this, instead of focusing on market share percentages, antitrust authorities began to focus primarily on consumer harm (which is much more challenging to show in court e.g. for predatory pricing, you cannot only argue that your competitor has very low prices right now, you also have to prove that they intend to raise prices once they achieve monopoly power).

The Chicago School has a bad case of thinking with their assumptions, instead of thinking to check their assumptions.

Jurisprudence also changed on the topic since then. In particular, Robert Bork published the legal theory of consumer welfare in 1978. SCOTUS essentially adapted that wholesale into case law starting with Reiter v. Sototone.

Given the current makeup of the court, it seems extremely unlikely that the justices would overturn this precedent.

The biggest challenge for the antitrust lawyers in this case will be to outline consumer harm. In more traditional cases harm is measured in dollars. Here it will have to be measured in privacy, free speech, etc. And that’s a much harder case to make. I even wonder if it’s possible to argue that services for which you don’t pay (Fb, Insta, and Whatsapp) and that are not essential can be considered monopolistic.

IIRC this is exactly what MS claimed about bundling IE : it was tied into the kernel.

Yes that was a defense. Explorer.exe and IE were intimately tied together. Remember “active desktop”? You could make your Windows desktop background a web page. Crazy times...

I wouldn't mind that, actually. I could create my own dashboard/background.

Well as always, that can easily be done on Linux. There is even a widget on KDE that lets you do precisely that.

Please say more!

I know of that: https://store.kde.org/p/998902/

You can also make nice updating (but not interactive) background with conky.

With Wallpaper Engine or similar programs you can still do it. It's no longer a Windows feature, but they have enough API surface to make it possible for other software to implement.

Bare in mind back then was before "Web 2.0", there wasn't AJAX, responsive web design and any of the other technologies we now take for granted. And to add to the woes, most computers still weren't powerful enough to handle running Active Desktop. It was a huge resource drain.

Some of the other desktop integration features of Internet Explorer 4 were pretty nice though.

These days most desktop environments allow for desktop widgets and usually there's a web view widget amongst them. So you can still have your web desktop if you wanted. At least now computers are powerful enough, and the frontend tools are useful enough, that there is some arguable benefit.

This can still be done.

And I loved it!

I don't think MS said it was bundled in the kernel, but it is true that parts of IE are tied into the OS (or at least used to be) in the form of mshtml and jscript.

Basic functionality like windows update was implemented using websites and activex components. You couldn't run that with any other browser.

Actually you could, as long as that browser used the Trident engine. Microsoft could have easily shipped a dedicated update application that's just a simple "browser" locked to the windows update website.

To be fair IE crashing would take out the whole OS.

Back then anything crashing usually took down the whole OS

Didn't MS still argue they couldn't unbundle IE in the XP days?

It's a lame excuse anyway. They could have just hidden the IE symbol, deactivating auto-open with IE and it would have been fine.

I think the arguments will be

- FTC Approved these as as legitimate acquisitions

- Facebook has direct competitors that can overtake them in just a few years

- Instagram was made by a few python coders in months. This shows how easy it is to make a product that competes with Facebook

> Facebook has direct competitors that can overtake them in just a few years

>Instagram was made by a few python coders in months. This shows how easy it is to make a product that competes with Facebook

Then why did Facebook buy Instagram, instead whipping up some python code on their own if it's so easy?

I would conclude they bought them to be anti-competitive.

For their users.

At that point (around the time of the acquisition), I guess one could say there was significant overlap between FB and IG users, and that FB had more users than IG.

No, Instagram had a thriving mobile user base. Facebook was getting demolished by Twitter in this area at the time. FB knew the future was mobile

That clearly implies that the acquisition wasn't for the users but for the platform and the brand value associated with being present on mobile.

Parts of IG in the early days involved filters on iOS, which at the time there was no library for. My understanding. The web version was Django, yes.

By web version I'm guessing you mean server side?

Server side was a few python components with lots of tricks in Postgres for scaling purposes. You didn't have any real choices for what language you used in iOS

Yes, that's what I meant. The point I was trying to make was that just knowing server side scripting was a necessary but not sufficient condition to create a hit photo sharing app at that time, because there weren't many options on the camera at the time, and the filters were what drove the growth at that time.

Agreed.

FYI that's exactly what they did when they broke up Ma Bell into the 7 RBOC's. Same company, same massive infrastructure. They owned everything and the government was still able to break them up.

What? How is that proof for a monopoly. Doubt things are that integrated but if they were then it would be proof of incompetence.

the funny thing is that this actually proves the opposite. The consumer overall would obviously be way better off with three redundant and independent systems that don't go down at the same time.

"Just like many independent companies use the same computing infrastructure (AWS) the three unrelated services can be spun off and pay FB Infra to use the messaging platform!"