Comment by AnthonyMouse | Hacker News Reader

Comment by AnthonyMouse | Hacker News Reader

Parent Full thread

What you're essentially asking for is the ability of someone to create an account on a service that isn't Facebook, but still be added as a friend by someone on Facebook, view content only available to their friends, share content with Facebook users, send and receive DMs etc.

That has the potential to be extremely effective. Anyone using a service solely for the network effect could switch to another one, which could be a large fraction of the users and reintroduce meaningful competition.

The biggest problem is protocol complexity. If Facebook adds some kind of timeline feature or anything like that, do they have to add it to the protocol other providers interface with Facebook using? If so, they can stamp out competition by adding so much complexity to the protocol that challengers can't keep up with it. If not, they can add features that challengers can't interface with. And expecting technologically-sideways regulators to arbitrate that sort of thing is a big ask.

I think some sort of framework to take your social graph and personal data with you and messaging as a baseline would already be a strong thing.

That's really the biggest issue when it comes to switching services, keeping your contacts and primary communication intact.

Without interoperability and universal identity, social graph doesn’t help.

Universal identity, you can maybe solve with email. But there are problems like requiring using real name, multiple accounts etc

Interoperability between messengers will be limited to maybe text and images. Even on this there are huge problems, imagine snapchat vs fb.

Isn't this what happened with Microsoft network protocols, court ordered them to be made public?

Does it benefit Facebook in any way to have protocol features that can’t be interface with? Why wouldn’t that result in non-consumption?

Sounds like a fun project.

>Why wouldn’t that result in non-consumption?

Because FB consumption happens 99% on their own webpage and mobile client.

Cell service providers had to do it. It wasn't that long ago in the US when you couldn't take your cell phone number to a different provider.

Facebook, et al, could provide restricted APIs that let another authorized service gain access to your information via your user identifier. I imagine they provide partial access for favored partners already.

Cell service is a commodity, though. They effectively act as dumb pipes for sound and data, and although they would like very much to not be that anymore, that is the reason they have customers. Also, there aren't many value-adds to be had at the protocol level. Everything else was already fairly standardized, so adding the number porting stuff on top of that was comparatively easy.

What we're talking about here is pure software interop. There is no standard, and any standard that would be devised would become a quickly-outdated albatross around the neck of the developers. Imagine writing code to interface with someone else's crap API, except now you're required to do it by law, and any new feature you come up with has to be either shoehorned into it, or more likely, go through a long and bureaucratic approval process, or even more likely, simply never shipped.

Now pretend instead of Facebook, it's your instance of Mastodon that you want to hack on. HN doesn't have the requisite font face options for the magnitude of NO F'ING WAY that is my reaction to getting the government this deep into the nuts and bolts.

Only solution is a commodity identity and data provider. But it would never happen. Moderation, cost, legal issues are impossible to handle.

I am confident that an interface standard could be developed to allow some level of identity portability. Supporting such an interface would be rolled into the myriad other compliance requirements that tech service providers have to deal with these days. (A natural part of being in a maturing consumer facing industry.)

Cellphone carriers argued that number portability was too hard or too expensive.

"But cell phone companies argue counter that the idea of portability should be scrapped because customers aren’t demanding it. They argue that implementing portability, which involves retooling the software of each carrier’s computerized call-routing system, would be expensive and technically complicated.

“The cost to be able to move numbers from one carrier to the other is significant, and it would be significant for the customer too,” said John Scott, deputy general counsel for Verizon Wireless. “Our argument is that it’s not necessary to promote competition.”

Carriers Aim to Kill Number Portability - 2002

https://www.latimes.com/archives/la-xpm-2002-jan-16-fi-cell1...

They did, but that's a pretty obvious pretext because anything that makes it easier for their customers to switch carriers is bad for business.

This is materially different.

Reply on news.ycombinator.com