Not everyone has access to a $100 server, reliable power, and unmetered internet, let alone the knowledge and time to be their own sysadmin.
What if organizations in a position to do so, from mobile phone operators to ISPs to public libraries, provided some basic level of content hosting with standardized content management APIs? People with the means and motivation could host their own, but most people wouldn't have to. Perhaps that could pave the way for universal front-end tools and independence from the social media giants?
Related: It looks like someone is trying to address the issue by layering a social media experience atop an existing distributed messaging system (Matrix): https://www.kombuchaprivacy.com/circles/
Then for actual social networks, things like briar or manyverse are perfectly functional and only lack network effect.
For actually community owned protocols (an example is matrix or fediverse), someone has to go first. May as well be you.
Make an account, ask a community group to be contactable without signing over your private info, get a couple of friends to do the same.
Recently I came upon the the phrase "cloud repatriation", which seemed a fresh angle on a word that's fallen on hard times.
Evacuate! Yes we need a Dunkirk for those helpless souls left on the beaches of Facebook. The idea that they're going to swim, one by one, back the safety of personal web pages is silly.
(Okay, okay, maybe a bit more work on security, but don't go so far as to make it bland.)
Maybe I my lenses are rose tinted, but that era of social media was fun.
I don't feel like I'm missing out on anything but the usefulness of the web lessens year over year. Now that the platforms have locked down even viewing an instagram or twitter fee without an account I almost wish I could just get rid of all links pointing to them.
The easiest thing would be to give in an make an account but I'm too stubborn ha.
Good riddance.
Come up with a way for anyone to publish content from their phone, with no subscription, and no need for an always-on server, and you'll have the basis for something less centralized.
And at the end of the day, there has to be some kind of server somewhere. Even if it's using SMS or DNS or something to accomplish this.
IE have a DNS where anyone can register whatever they like. That way you can send a message direct to your friend's phone with no middleman.
My invariably naïve thoughts on the SuperSimpleSecretsServer:
- Alice generates 2 publicPrivate key pairs. 1 for the SSSS and 2 for their own use.
- Alice contacts SSSS with PublicKey1.
- SSSS responds with randomstring.
- Alice privatekey1 encrypts the random string and privateKey2 encrypts her IPv6 address and sends both to SSSS.
- SSSS authenticates and stores the encrypted-IPv6-blob against that publickey.
Bob does same.
Alice and Bob exchange (PublicKey1 and PublicKey2) however they do. Alice and Bob can now query SSSS for IPv6 of the respective key(aka person they want to talk to) whenever they want.
But I honestly don't know much about crypto, security, scaling, routing, replication, who'd run it etc etc etc. I just plain don't know much....
1. the focus is still on human readable names and converting that to an IP.
2. the IP address seems plain text to anyone who requests it. Not a key encrypted blob.
So. Not really like what I was describing at all. Maybe it could be tweaked to behave like I wish and it could solve the other aspects of such a service already.
https://www.gnunet.org/en/video.html
The Videos are kind of hard to watch, so maybe the slides are enough to understand how it's supposed to work:
https://git.gnunet.org/presentations.git/plain/icann66/20191...
I am picturing the use case of such a service as different to how we use DNS today, but I think this would support it.
Merely keeping the current system, but wishing consumers would use it differently, is a utopian dream that is unlikely to come true. You need to look at why consumers behave the way in the current system, and then you have to imagine a different system, and that different system cannot look anything like IP/TCP/UDP.
We currently have like 8,000 RFCs. We've designed the world's most flexible system. It is very complex. I think you'd need a system simple enough that you could specify the whole thing in a 50 page document. That kind of simplicity. The simplicity should allow it to be cheap, and therefore easy to subsidize, either by the government or out of the profits of cable or phone or network companies.
You can pay someone for that, somewhere around the order of $5 a month, probably less, or free but ad-supported.
And that's more or less what Facebook is, except we pay with our data and attention on the back end rather than our dollars up front. The product might be rather different if we did pay in dollars, and for that reason I suspect digital protocols aren't the fundamental problem.
$5 a month is far too expensive for most of the world, which is why so much of humanity is now in the walled-gardens like Facebook. So if we want to find a way to destroy the walled-gardens, we're going to have to find an approach that is orders of magnitude cheaper than $5 a month.
Any amount will always be a sacrifice compared to free, but given that "free" actually has a cost in terms of privacy, attention, and even baseline function of the tool, there's really no free lunch here, just the currency you want to pay in. Making it actual currency lines up certain incentives better. Even if it doesn't solve the problem among, say, food insecure.
Why? You're assuming your conclusion. A different protocol can have different starting assumptions.
Then there's fediverse or even briar or manyverse that work without internet.
translate into (roughly): "network protocols are too complex, I want simple text only"?
I just don't see how those are related, why we'd have to restrict ourselves to text, why IP is somehow not "enough", etc.
why IP is somehow not "enough"
How did you get that from my comment? I said we have almost 8,000 RFCs, IP is very robust, but in its complexity there is a certain minimum, unavoidable expense. If we're looking for a protocol that is cheaper, and therefore easier to subsidize, I'd guess that we need a spec so simple that you could write it out in a 50 page document.
But the idea that we can't have a grassroots socially networked connection of computers on the internet because of the protocols? I just don't see how you've made a case for that. I certainly don't see why it'd have to be text only.
The complexity of the internet has grown organically based on responding to limitations (mostly). No doubt there's some unnecessary complexity in there, but what you're talking about is essentially throwing everything away. I just don't see how you've made a compelling case for such an expensive proposition.
We need something cheaper than what we have now. People gravitate to walled gardens like Facebook because setting up their own systems is too expensive.
"No doubt there's some unnecessary complexity in there"
At no point did I mention the unnecessary complexity of IP/TCP/UDP, I'm not sure where you got that. I'm only talking about the necessary complexity. It's a great protocol that enables the modern Web. But it is complex enough that there is some expense to setting everything up, and that is why humanity ends up in walled gardens like Facebook. If you want to destroy the walled gardens, then you will need a simpler, cheaper protocol.