Allow me to digress from the usual topic on this channel once more.

I'm pretty sure that no human being on this planet has created nearly as many federated social platforms as @mike. But all these (actually not always so) different platforms can be a bit confusing. Even I may be wrong here and there, but I'll try to make some sense of them by putting them into a kind of chronology.


So first, there was #Friendica. Only that it started out under the name of #Mistpark. I'll get to the name later.

Remember #Diaspora? Remember summer 2010 when the crowdfunding run was launched so that those four guys could spend all their time creating a free, #OpenSource, decentralised, federated social network (a.k.a. #Facebook killer) which they wanted to name Diaspora*?

Well, they unknowingly wanted to re-invent the wheel. #StatusNet was already there, #GNUsocial was already there, and especially, Mistpark was already there with a 1.x release and more powerful than both, actually, more powerful than Diaspora would ever become. I think Mistpark even already had Diaspora*'s aspects, only that they were called groups.

As for its concept, Mistpark went beyond that of Diaspora*. Mistpark didn't only want a bunch of instances ("nodes" in this case) of its own kind to connect with one another, it also wanted to federate with everything else that moved, be it e-mail, be it StatusNet, be it Twitter, be it whatever.

The first name change was from Mistpark to #Friendika. The reason was that the original name sounded repelling to German speakers. "Mist" means "fog" in English, but "dung" or "manure" in German, not to mention that it's a German curse word.

When Diaspora* was finally there, Friendika didn't see it as competition, it saw it as another federation target. To this day, Friendica is fully federated with Diaspora*, and that has exclusively been the work of the Friendika developers who studied Diaspora*'s source code and reverse-engineer it because it didn't have an API.

Probably the biggest coup was the bidirectional federation with Facebook. This was what everyone was waiting for. This, however, was also where the trouble started. Facebook didn't want to be federated with a non-commercial social network and started taking defensive measures. Also, Friendica users (the second name change was through meanwhile) who used the Facebook connector had their entire and often very busy Facebook timelines mirrored onto Friendica nodes, one of the reasons why even nodes on powerful root servers often had to close new registrations even though they only had a little over a hundred users. So there were several reasons why Facebook federation was axed again.

Internally, Friendica uses its own protocol named #DFRN. But I guess Mike had meanwhile seen it as a dead end, also because he had a new idea: #NomadicIdentity, not only the ability to easily take your account from one instance to another, but the possibility to have it on multiple instances at the same time, keeping the copies in sync.

That's why he laid the foundation for a new protocol that could do that: #Zot.


And with it came the next social platform. It was first just simply named Red from Spanish "red" = "net". Red was based on Zot from the beginning, and as an experimental platform, it only understood Zot. On Friendica which was now running at full steam on dozens upon dozens of nodes, and which Mike had passed on to the community, the development was followed with interest. And just like later platforms, I think Red actually got a few small public instances because someone really wanted to try it out. Red eventually changed its name to #RedMatrix.

Also, Red didn't just want to be a social network like Friendica. The idea was rather to have a "social content management system" that could do just about everything you could do with a website and/or a cloud server. Third-party federation was slightly reduced, connections to commercial platforms didn't come back. But as Red evolved, the Diaspora* connector was included which was also used to federate Red with Friendica.


From the Red Matrix emerged #Hubzilla, the Swiss Army knife of the #Fediverse. Still today, its possibilities have rarely ever been fleshed out: not only microblogging, but macroblogging, article publication, websites, wikis (no, I'm not kidding), #WebDAV, #CalDAV and #CardDAV server and so forth.

Next to the nomadic identity that came with Zot, Hubzilla introduced another killer feature: one account, many separate channels. Each one of these channels is basically like one Friendica account. You can have multiple fully separate identities on one account, and nobody (except the instance admin) can tell that they're all you. So this goes way beyond Friendica's multiple profiles. By the way, Hubzilla still has multiple profiles per channel.

Some say that the Red Matrix was renamed Hubzilla. This isn't true. Hubzilla is a fork of the Red Matrix, one could say it was a stable snapshot of the Red Matrix.

For the development of the Red Matrix continued. Planned advancements on Zot couldn't be tested on stable Hubzilla, they needed their own testbed. Eventually, the last Red Matrix instance was Mike's personal one with himself as the only user. It still federated with Friendica and, of course, Hubzilla.

In the meantime, #ActivityPub came along. It wasn't just another obscure networking protocol, though, because #Mastodon made it huge. So at least Friendica and Hubzilla had to adopt it. Friendica firmly integrated it. Hubzilla made it into an app just like all other protocols that aren't Zot because they stand in the way of fully nomadic identity. By the way, both profited from its introduction because the federation between each other no longer had to use the Diaspora* protocol.


For the next advancements of Zot, two new platforms were forked from the Red Matrix or Hubzilla. At this point, Mike wasn't involved with Hubzilla anymore either. First, there was #Osada, an early testbed for what would become #Zot6, but still with ActivityPub. For pure Zot6, #Zap followed suit. Most connectors that are neither Zot nor ActivityPub, including the one to Diaspora*, weren't taken over, as were many of Hubzilla's extra abilities (websites, articles, wiki, CardDAV, two parallel calendar systems etc.) to keep it slim. It did get to keep the various types of channels as well as one CalDAV server and the WebDAV connection, though.

Eventually, when Mike handed them over to the community, they used the exact same code base. The only differences between Osada and Zap was whether or not the admin had ActivityPub on (Osada) or off (Zap) and the name.

As having two different names for the same thing, depending on the instance configuration, Osada was discontinued in favour of Zap which now included ActivityPub itself. In the meantime, Zot6 became stable and was backported into Hubzilla which thereby became fully compatible to Zap, only that what Hubzilla can that Zap can't cannot be mirrored to Zap.

Then Osada re-emerged as Zap's unstable branch. Along with it came a new Red Matrix which, as far as I could see, was now an even more purist, even more unstable branch that only served for testing Zot8 and lacked all other protocols.

To top this off, in 2020, Zap itself got a stable branch even more intended for productive use. For this purpose, the name Mistpark was dusted off. The new stable branch was named #Mistpark2020 or simply #Misty. Misty was the first of its kind to not even get an announcement anymore, though. Its home page on Zotlabs disappeared along with Zotlabs before it could be filled with any useful information.

Two things were interesting: Red Matrix, Osada, Zap and Misty were based on various states of the same code base. It was possible to switch from one to another by rebasing the local code repository on your server. This became obvious through instances that carry the name of one project but run another one.


It must have been in 2021 when #Roadhouse showed up, again, unannounced. It seemed to be nothing more than a concept for the next generation of distributed social platforms. Roadhouse was the first of its kind to use the #Nomad protocol which, I guess, is forked from #Zot because it serves the same purpose. It got its own home page on Zotlabs which remained as uninformational as Misty's.


And then the most recent name popped up: #Streams. At first, it was even less clear what Streams was supposed to be and what set it apart from Roadhouse, not to mention Red Matrix, Osada, Zap and Misty, also because Zotlabs didn't say what Streams was either.

But I guess Streams' purpose has emerged in the meantime through word-of-mouth: It's the experimental successor of all five and the solution to this maze of names. Streams isn't even a product with a name, it's a concept that uses Nomad for nomadic identity and that is in constant flux, hence Streams. The idea was to do away with fixed names to get rid of the previous chaos. Everyone can name whatever they do with Streams however they want.

There is currently only one more or less public Streams instance, but it still carries "Stream" in its name. At least two more instances which may be private are named something with "Streams", too. So whether Mike wants or not, Streams has become a name of its own, and people use it.

How many Streams instances exactly exist right now is hard to tell, even from Communities pages on Streams instances or Sites pages on older platforms, because they don't necessarily identify themselves as Streams instances. So if you go through one of these pages, and there are names in the Projects column which you don't know as Fediverse platforms, check out what's behind them. It's often only one instance. Open the instance, click its burger menu, and if there's a Communities link, it's a Streams instance. I've discovered a lot of Streams instances not named anything with Streams this way. Private instances included, I guess Streams must have more than a dozen instances already.

There has even already been a request to launch a Streams support forum much like the one for Hubzilla; after all, Streams still supports forums. It's safe to say that Streams is doing quite well for something so obscure.

Feature-wise, Streams is the same as Zap and Misty.


But what became of the six platforms between Hubzilla and Streams?

• Red Matrix kept having only this one single-user instance because nobody else dared to touch it and set up another instance. It's a Zap instance now as far as I can see.
  
• Osada never really took off, Zap probably did only after Osada was merged into it, and some Osada instances became Zap instances. This explains why Zap has got comparably many instances. Most of them, however, are tiny, probably private and utterly undermaintained as they run rather old Zap versions. Zap only lives by numbers, and it's the only one of the five listed on Fediverse Observer. Also, while the FediDB lists all five, it only knows that one Dominican public Zap instance and none of the others (looking through its connected sites reveals many unlisted instances of Zot-based networks, by the way). Still, it seems to be on the deathbed, being superseded by Streams, experimental as the latter may be.
  There still seem to be a very few running Osada instances, but Osada can be considered dead as the focus is on Streams now.
  
• Misty didn't take off either, even though it was considered more stable and more production-grade than Zap. This time, the reason may simply be because Misty got zero advertising, so nobody heard about it, probably not even some of the Zap crowd. Misty never had many instances, they weren't properly advertised either (the same applies to most Zap instances, by the way), and Misty's death knell may have been the unannounced shutdown of its largest instance. Basically, there was little room for Misty next to less obscure Zap.
  
• Roadhouse didn't even manage to get much limelight before Streams appeared shortly afterwards. In both cases, the only way to find out what they were and what they did was by either studying the source code or installing a private instance. Streams, however, had the advantage of being even newer. The-Federation.info knows exactly one German Roadhouse instance which was originally set up as Misty and has meanwhile been upgraded beyond Roadhouse to Streams, and there only seems to be one remaining unlisted Roadhouse instance.
  
• I've seen another result of an upgrade from Zap to Misty. So it's safe to assume that you can upgrade all five to Streams. If this is the case, then now that Streams is here, it probably isn't worth spreading the developer community across six almost identical platforms. Basically, Streams has become the latest version of Red Matrix, Osada, Zap, Misty and Roadhouse.
  
• At least Red Matrix, Osada, Zap and Misty are still being maintained in a sense, though. All four got the same small Git commit from Mike a good month ago. Roadhouse got one four months ago.

As of now, Friendica is still going strong, so is Hubzilla, and Streams seems to be cleaning up the mess that came after Hubzilla.

If you really want to try out something with Zot, my current recommendation is Hubzilla, even if it may seem bloated and cumbersome to you, even if you'll never harness its full power. Many of its extra functions are additional apps and switched off by default; this includes ActivityPub, by the way, this is important to know.

It's hard to find a public Streams instance with open registrations currently, much less multiple ones that'd be required for a nomadic identity. Neither Fediverse.party nor the FediDB nor The-Federation.info nor Fediverse.info even knows Streams, and existing Streams instances usually don't identify to other Fediverse servers as Streams instances. It's still a rather underground and grass-roots project with no publicity at all. As Streams is rather experimental, however, you may want a nomadic home on at least two instances to have an instant backup, should one of them shut down.

Zap has got exactly one instance open to the public, and seeing as Zap may be shrinking rather than growing, I don't expect this to change. Again, due to Zap's still small size and unclear future, I wouldn't recommend using it without nomadic identity as a safety net.

As for Osada or Misty, good luck finding an instance to join, much less one that's here to stay and ideally be upgraded to Streams one day.

Hubzilla may not be as bleeding-edge as Streams, and it may be overkill for your purposes if Zap or Streams would be sufficient, but it's stable, it's big enough, it's established, and it's different enough from Streams to not be endangered by it. I mean, Hubzilla hasn't managed to kill off Friendica either, right?

I've just written about how Second Life is largely ignored in the Metaverse context, but it's much much worse with worlds based on #OpenSimulator. Next to nobody even knows that it exists beyond its users, its former users and a few people in #SecondLife.

Why does it matter then? First of all, because #OpenSim is the technical foundation for something that comes as close to the literal #Metaverse as we've gotten thus far: the #Hypergrid.

Unlike just about all the other #VirtualWorlds, OpenSim isn't one monolithic entity completely owned and operated by one company or foundation or so. It isn't a virtual world per se. The name rather stands for thousands of big and small virtual worlds, so-called "grids". The world of Second Life is being referred to as a "grid", too. All these worlds are organised in squares of 256x256m, that's why.

In the case of OpenSim, these grids aren't just places within the same world. They are actually separate worlds. Each one of them is operated by someone else, and each one of them is even hosted individually, some on rented webspace, some on machines the grid owners run at home. Some are owned by companies, some by foundations, most by private persons. And they didn't "rent these worlds from OpenSim", they created them from scratch. There are no higher powers within the OpenSim ecosystem than the grid owners. It's the same as with e-mail or #XMPP or #Matrix or #Mastodon or any other project on the #Fediverse.

And here is where the #Hypergrid comes into play: At least 95% of these OpenSim grids are connected with each other, much like e-mail servers or XMPP servers or Mastodon instances or #Diaspora pods or #Friendica nodes or #Hubzilla hubs. Applied to virtual worlds, however, this borders on a sensation: You can have an avatar on one grid, and you can travel to all other grids on the Hypergrid. You usually even take your entire inventory with you, you can even pick up things on other grids and take them home with you, and you can become friends with avatars from other grids.

Essentially, what some big corporations and start-ups try to create from scratch right now has been around and in use since 2008 already. Not experimental use, but everyday production use. For something with so few developers behind it who, due to OpenSim's decentral nature, have to take care of various products, namely OpenSim itself and the viewers, this is remarkable.

In other points, OpenSim is very similar to Second Life. Community-building and creative possibilities are largely the same, only that creativity is stifled by the massive influx of high-quality payware that was stolen from Second Life and is now offered as freebies all over the Hypergrid. I'll come back to that. Okay, and because OpenSim is dirt-cheap in general, and because you can get far without paying a penny, most OpenSim users are unwilling to spend any money on it.

They also laugh about outrageouly expensive land on other worlds. Many of them sell land for hundreds of thousands or even millions of US dollars. Second Life offers land rentals; for example, a 256x256m region on the Mainland costs over $300 a month. Many OpenSim grids offer similar land rentals, but you can get a region of the same size for usually under $20, mostly under $15, often under $10 a month. Or you can host your land yourself and attach it to a grid. Or you can even run your own grid and attach it to the Hypergrid. Not only do most grids have such vast coordinate ranges that they're unlikely to run out of land anytime soon, but you can literally create your own new land.

Of course, all this is a nightmare for those who want to cash in on land sales or land rentals. And the abundance of land leads to the Hypergrid feeling even more like a wasteland than Second Life. By far most people own at least one region, many own more than one, and my estimation is that one out of a dozen OpenSim users is a grid admin.

But why does nobody know OpenSim? Well, I'd say that a key reason is that it's too decentralised. It doesn't have an umbrella organisation that a) represents OpenSim to the outside world and b) at least tries to steer what's happening in the ecosystem. OpenSim itself, i.e. the server application, is "represented" by, I think, one single remaining permanent developer, and even Ubit Umarov only works on OpenSim in his spare time. He doesn't even have a grid of his own, only a private region on OSgrid.

This means that there is no "official" grid. OSgrid comes closest. It was the first public grid, it's the oldest grid, it's the biggest grid (it has more land than Second Life), and it's still the grid with the most users. But it is not official. And its only connection to the OpenSim dev team (if there happens to be one) is that it contributes a lot of code. After all, OSgrid is still the same experimental, bleeding-edge grid that it was in 2007 when it was launched to test new features in public.

Like most other big grids, OSgrid presents itself as a stand-alone virtual world. It doesn't even mention that other worlds like it exist. It does mention the Hypergrid, but instead of explaining what the Hypergrid is, it only gives Hypergrid addresses to its most important official places. It also has its own forums that seem to try their best to blank out the existence of worlds outside OSgrid. It actually isn't too uncommon for grids to have their own forums. There is, however, not a single general and grid-independent OpenSim forum in English.

Mutual acknowledgement or even collaboration only exists between smaller grids that are less or not at all interested in competition between grids. This includes grids that rely on visitors from the Hypergrid because they've decided not to have a big community of their own, sometimes limiting the number of residents to the grid owners.

OpenSim also doesn't have an official viewer. There are several viewer projects, usually third-party Second Life viewers that also work with OpenSim, but each of them is its own little bubble with hardly any connection to elsewhere. In fact, the most popular viewer, the Phoenix Firestorm Viewer, can barely keep its OpenSim-specific development running and has to focus on Second Life.

The official OpenSimulator website, and it is actually official, is a wiki with operating a grid as the main focus and grid software development as the secondary focus. Its end user documentation is scarce, even less maintained than the rest of the wiki and thus not as up-to-date as it should be. Especially the list of public grids, basically the only connection between this website and the rest of the ecosystem, tends to be outdated. The Second Life Knowledgebase is more useful in anything that isn't OpenSim-specific, and for OpenSim-specific things, you often have to ask for help in-world.

OpenSim's central hub is OpenSimWorld which is "third-party" again and neither affiliated with the OpenSim devs nor with any one grid. It hosts the closest thing to a central community, but unfortunately with little to no moderation because the one-man staff sees himself as nothing more than the tech admin of a sim catalogue. OpenSimWorld also rendered in-world teleporters to other grids (and even within a grid) largely "obsolete", so these connections between grids and mutual acknowledgements are getting rarer and rarer.

All this contributes to OpenSim being basically completely unknown. Nobody feels responsible enough for OpenSim to advertise the ecosystem as a whole to the outside world. Everyone only cares for their project. The dev team is only that and only responsible for the server application. They would never advertise anything beyond that. Grids only ever advertise themselves, if at all, and even then mostly within the OpenSim ecosystem. If a grid significantly advertises itself beyond that, not only does it do so as a stand-alone world like Second Life or #HorizonWorlds, but grids that do so are usually grids that aren't really trustworthy. There seems to be more competition than cooperation between grids, not to mention bickering and long-standing petty feuds.

I think it'd be worth representing the OpenSim ecosystem at #FLOSS or hacker events so that the #FreeSoftware and #OpenSource scene becomes aware of it. But the sad fact is that there's nobody who could do that. The ecosystem as a whole doesn't have any representatives, nor does it have anyone who could appoint official representatives, and if you sent someone from a big grid, they'd probably advertise their grid first and foremost and give the impression that their grid is OpenSim.

Lastly, word of mouth doesn't work either because hardly anyone ever talks about OpenSim outside OpenSim.

Within all the ruckus about the #Metaverse, this is more than a pity. Companies and others try to create something they believe they'll be pioneers at, not knowing it already exists. It's quite similar to whenever someone decides to develop a free, decentral, distributed social network, not knowing that these have been around since #GNUsocial and Friendica, formerly known as #Mistpark, because they, too, get next to zero publicity.

It's also a pity because the developers and designers of the coming Metaverse could learn a lot from OpenSim. They could look at how worlds are connected with each other over the Hypergrid. How friendships and groups work on the Hypergrid, if they do, and if they don't, why. They could learn about enforcing rules, guidelines and content ratings in an ecosystem that isn't ruled over by one entity at the top or rather why certain Second Life approaches were bound to fail in OpenSim.

And they could gather some experiences in content protection, especially when, how and why it fails in Second Life and even much more so in OpenSim. Even better if they come to their senses and see that #NFTs on a #blockchain aren't a sensible solution because they've got their own ways of abuse which are being exploited right now.

There's the so-called Metaverse Standards Forum which tries to develop what OpenSim has been running for 14 years now. OpenSim, however, isn't represented in the #MetaverseStandardsForum at all, not by anyone. Because OpenSim doesn't have anyone to represent the entire ecosystem. Because nobody who deals with OpenSim feels even remotely responsible for the whole thing. Not the core devs, not a single one of the grids, not OpenSimWorld, not the Infinite Metaverse Alliance, nobody.

Everyone just cares for their niche. You could have half a dozen grids join the MSF, but then you'd end up with six grids representing themselves as individual grids rather than as parts of something bigger and overarching, and you still wouldn't have the OpenSim core dev team on-board. The core dev team, in turn, can barely keep up with developing OpenSim, so how should they have time to represent OpenSim in the MSF?

And so, the MSF will continue creating something which they wrongly believe has never been done before because there's nobody there to tell them about OpenSim. They'll have to design from scratch, from zero what they don't know already exists with open source codes to look at and with 14 years of experience to learn from.

By the way: #LindenLab isn't a member of the MSF either.

If someone actually started an umbrella organisation for OpenSim, and be it only to represent it, that might help, but it's also likely that it wouldn't because it would only join a tiny fraction of all possible participants under its roof. Important potential members wouldn't have enough time and/or energy, and I'm sure this will include the core dev team. Others simply won't be interested in contributing to anything beyond their niche. Others again will boycott it if they see the "wrong" grids or people having joined it. They don't want to collaborate with "this" grid or "that" person, and be it due to some six-year-old personal beef. Yet others will refuse to join because they dislike (one of) the founder(s), or because (one of) the founder(s) is a resident on the wrong grid. Of what few members the organisation may have, some will try their best to misuse their membership only for their own profit while others will quit again due to too much drama.

And it'd be everyone for themselves again like it is now.