Diaspora was originally conceived as "a Facebook killer".

I love how people object to Bluesky being a commercial enterprise, even if it is a registered "public benefit corporation". My question to those people is, was your car built by a commercial enterprise? How about your house or your bicycle? Do you get your groceries from a commercial enterprise? How about your cell phone or the phone's service? I haven't seen a non-profit, community-owned car manufacturer yet!

There is not one example of a social network that is a non-commercial enterprise that can be considered anything close to a success today. D* is the poster child for that: 15 years on the internet and 1,500 active users. And most of that lack of success is due the fact that anyone in the community can start a pod and run it. They do, people join and then the pod owner loses interest, runs of money or similar, and it shuts down, taking all its users off the network. Sure you can start over on another pod, but 90% of users on a closed pod don't, they just quit. The non-profit, distributed social network experiment has not been a success and specifically because of its non-profit, distributed nature. After being here on D* since almost the very beginning I am personally will to give Bluesky's "public benefit corporation" model a chance, especially since you don't have to quit being here to be there as well.

Adam, what has kept you on Diaspora all these years?

A good question! Mostly the interesting exchange of ideas and conversations around here, plus #caturday!

I think D* is just about an ideal social network in terms of engagement and the way it works. There are some outstanding "feature requests", as Jodi noted above, but overall it works really well. As I noted above the biggest problem is its "community-run disbursed network model" which is what I think has been the biggest reason for its lack of success.

By the way, the management at Bluesky, including the CEO, Jay Graber are experienced devs with strong backgrounds in how previous social media projects have failed. Not sure they will fall into the same old traps. In fact in interviews Graber has identified those traps and indicated how they intend to avoid them. Right now the plan is to not sell advertising to put into people's feeds and instead come up with a subscription model that will provide the basic features for free and some premium features for a price. Running a social network costs money, it can never be all free-of-cost, even if just for the server time, so it has to have some income. Even here on D* we all make donations to cover the costs and volunteer time to make it all run and work. I would be happy paying a monthly fee or even making a regular donation to keep Bluesky working. The users do not have to be the product if there is no advertising.

This is the biggest reason I am here. With no central point of control and no central place to easily collect data, this is the perfect design for communication.

LOL! No it isn't. In the D* model every pod shares all data with every other pod. If you wanted to scrape all the data, public and "limited" on the network all you have to do is open a new pod. You wouldn't even need to open it for sign-ups and when it synchs you will have everything everyone has ever posted for whatever purposes you like. For instance here on diasp.org we still have all the posts available from joindispaora.com, even though that pod closed years ago. With a commercial provider you at least have some data security, unless they get their whole database hacked. On a disbursed network you have none at all. It wouldn't work if you did. Note too, when we say anyone can open a pod that means "anyone". There is no sign-up security clearance. Anyone can open a pod: Somali pirates, Russian state hackers, ISIS, anyone. Later, once we figure out what they are up to they can be disconnected one pod at a time, essentially a pod-level block, but if data collection if your fear it is long gone by then.

Are you saying that there are not enough servers?

To handle 1,500 active users? One laptop would be enough for that. D* actually has lots of servers, one per pod, but someone has to pay for buying or renting those, plus running and maintaining them. It all costs money and it is all covered by someone for each pod, usually by donations.

I’m wondering where the “Most Free” pods may be.

No idea, really. It all depends mostly on who is running them and what you mean by "free". We have had pods in places like Russia that never responded to any outside communications, but they seem to have mostly closed suddenly. I am sure in the ISIS case any pods that didn't take them down were disconnected from the rest just due to fears of promoting terrorism.

If you have pod admins who don't communicate with anyone or take any action on their problem users than perhaps that is "most free", but it also means the rest of the network has to deal with spammers and scammers from that pod, inevitably. It is like on "X" where "free speech" means maximum Nazis.

The overall point is that the 15 years of experiences with a disbursed network here shows it is not a good model, mostly due to its inherent unreliability. It would work better if it was run by a single organization underpinning it, whether a non-profit or a "public benefit corp" that would prevent pods shutting down. Right now we are probably one pod away from D* being virtually gone for good and that one pod is being maintained and run by one person with no back-up.

Right now we are probably one pod away from D* being virtually gone for good and that one pod is being maintained and run by one person with no back-up.

This isn't a problem with the distributed design, but a failure to maintain it. Sadly, centralized, for-profit designs are inherently more supportable - once they reach a critical mass.

As for the "unreliability", Diaspora has been in operation for over 12 years, which is a pretty good record.

If you have pod admins who don’t communicate with anyone or take any action on their problem users than perhaps that is “most free”, but it also means the rest of the network has to deal with spammers and scammers from that pod, inevitably.

You are confounding two or three different things here. Pod admins who "do not communicate" are not necessarily not dealing with problem users. Also, admins who "communicate" are not necessarily dealing well with problem users. From what I see, or maybe rather believe, what we need are more pods, and especially more pods which are more free from government and commercial pressure.

Perhaps there should be a campaign to encourage (somehow) the opening of more pods?

Last time I was able to see the stats we actually had more pods than we have ussers, so that is not the issue. In most cases each pod is started by one person and run by them. Setting up and running D* pods is not a trivial task, it is complex and requires specific skills. Most pods have no back up person to take over, as in the Pluspora case, if they become "unavailable", once the money runs out for the rented server or the power gets cut off (basement server) that pod is gone forever. In military terms we would say that the is no "defence in depth". The pods are a single thin line and easy for holes to appear. The answer is not more pods all run by one person, it would be pods run by organizations with some serious redundancy and organizational depth. That way if one person dies there will be someone to take over and keep it going.

My experience in writing to pod owners who have problem accounts on their pods are that if you don't hear from them the problems are not being handled there quietly either. There have been pods on-line that you can tell no one is watching. The software often gets out of date, no updates, no communications, nothing, just humming along in the dark. Some of those have had to be disconnected from the network just due to old out of date software.

This isn’t a problem with the distributed design, but a failure to maintain it.

No it is a inherent and unintended design flaw in the concept. The fact that anyone can start a pod means that lots will appear and then disappear when interest wanes, the person dies or the money runs out. It wouldn't matter, except we lose most of the users on that pod, because, as I said, most will not bother to to try again on another pod. If you restricted pod ownership to organizations with the depth to reliably run a pod that would change the whole game and it might have grown the user base, but where is the incentive to do that? It would have to be run on donations to pay for itself. It has just proven to be a model that barely works, at least enough to keep it alive, but a long way from a success.