It isn’t reinvention. It’s evolution.

In 2011, Mike invented the Zot protocol because he couldn’t implement nomadic identity in Friendika’s DFRN.

In 2012, Mike forked Friendica to Red (later Red Matrix) because he couldn’t replace DFRN with Zot on something that people relied upon as a stable daily driver.

In 2015, he turned the Red Matrix into Hubzilla because the Red Matrix was Friendica with nomadic identity. And nobody needed that because everyone and their dog hosted their own single-user Friendica nodes. So the target audience had to change, and thus, the software itself had to be expanded.

In 2018, he forked Hubzilla into Osada and Zap because he wanted to develop Zot6, and what he envisioned Zot6 to be like would mean a whole lot of breakage. Like, you couldn’t have both nomadic identity and non-nomadic protocols in the same software. That’s why he couldn’t develop Zot6 on something that, again, people relied upon as their stable daily driver. So he had to fork Osada (non-nomadic, but with support for ActivityPub) and Zap (nomadic, but only supporting Zot) off Hubzilla.

In 2019, it turned out that Zot6 does play along with non-nomadic protocols. So Mike discontinuned Osada, forked a new, nomadic Osada off Zap and added ActivityPub to it whereas Zap stayed Zot-only.

Over the course of 2019, Zap itself got ActivityPub support. Thus, Osada turned from “Zap, but with different branding and ActivityPub” to “Zap, but with different branding and ActivityPub on by default”. Otherwise, both had the same codebase. And so Osada was discontinued.

In 2020, Mike wanted to advance Zot even further. Zot6 was good enough to be backported to Hubzilla. But Zap, like now-defunct Osada, had been pronounced stable, so there was no tinkering with it. Something something people’s daily driver. Even though only Hubzilla users even knew that Zap existed, and only few of them were willing to switch because they saw Zap as “Hubzilla, but without diaspora* and without articles and without cards and without wikis and without webpages etc. etc. and with no clear advantages over the real deal”.

So Mike took Zap and created three more forks: another Osada, Mistpark 2020 (a.k.a. Misty), Redmatrix 2020. Don’t ask me what was forked from what.

Rumours had it that this was a case of different levels of stability vs bleeding edge. Allegedly, Zap was stable, Misty was testing, Osada was unstable with ActivityPub on by default so that its interaction/interference with Zot8 could be tested, Redmatrix was unstable with ActivityPub off by default so that it doesn’t stand in the way.

In reality, Osada, Misty and Redmatrix were identical in everything but branding. The reason why they were three was because Mike wanted to confuse the hell out of brand fetishists who used $FEDIVERSE_PROJECT out of nothing but brand worship with no regards for features. Like, people who refused to switch from Mastodon to clearly superior Zap because Mastodon was the cooler brand.

In early 2021, Roadhouse joined the fray because Mike wanted to go beyond Zot8, and people seemed to daily-drive Misty and Osada now. This time, all backwards compatibility was to be sacrificed. Thus, Zot11 wasn’t Zot11, but Nomad. Roadhouse had to have support for “Zot before Nomad” added, that’s how incompatible Nomad is with the old Zot.

At this point, Mike maintained five Fediverse projects:

• Zap
• Osada (III)
• Misty
• Redmatrix 2020
• Roadhouse

Enough to really confuse the brand fetishists.

In October, 2021, he really flipped them the bird when he forked Roadhouse again. This time, he fully intentionally removed any traces of a name, the branding, almost the entire nodeinfo code and its license. Unfortunately, while he could deprive the software of a name and a logo, he couldn’t do that with the code repository for which he chose the name “streams” and that logo with the three blue waves.

Now he still maintained the same five Fediverse projects plus that nameless thing that, according to him, isn’t a project. But that nameless thing was the only one out of the bunch that he actually developed. Everything else was in maintenance mode.

Came New Year’s Eve, 2022, and Mike put Zap, Osada, Misty, Redmatrix and Roadhouse on the chopping-block. They weren’t needed anymore. What people had started calling “(streams)” was now stable enough to replace all five.

From then on, Mike only worked on (streams) anymore.

In 2023, silverpill, the creator and developer of Mitra, was absolutely hell-bent on making Mitra nomadic. But he didn’t want to switch to Nomad. He wanted to do nomadic identity with ActivityPub. And so he hit Mike up, and the two started brainstorming about how to pull this off.

This time, Mike didn’t fork anything, even though, yes, people were daily-driving (streams) now. Instead, he created a “nomadic” branch of the streams repository to tinker around with implementing nomadic identity in nothing but ActivityPub.

Fast-forward to summer, 2024. Mike was so confident in the “nomadic” branch that he merged it into the “dev” branch. Soon afterwards, he merged the “dev” branch into the “release” branch. In doing so, he officially switched (streams) to decentralised IDs as per FEP-ef61 “Portable Objects”.

It. Blew. Up. Big. Time.

It had worked just fine under lab conditions with only Mike’s instances and silverpill’s non-public development instance of Mitra as sparrings partners. Out in the wild, it blew up. (streams) no longer properly federated with anything.

The reason: (streams) had to deal with so many IDs now that it confused them.

The consequence: Mike had to work his butt off trying to fix that mess and figure it out first, even though it was only for a handful of users.

In mid-August, he forked Forte from the streams repository. One of the first things he must have done was rip out any and all support for Nomad and Zot6 to get rid of at least some IDs.

Also, he attributed the lack of success for (streams) to the Mastodon-centric Fediverse rejecting something that’s no project with with no name, no brand and no license that doesn’t submit stats. Thus, Forte was declared a project, it got a name, it got a brand identity, it got its nodeinfo code back, and it got its MIT license back.

August 31st. Mike was so burned out from all this that he officially quit and retired from developing software. Effective September 1st, the streams repository and Forte were up for grabs. As there was no-one there to grab them, Mike still went on working on both, including introducing new features to both. After all, new code for Forte could fairly easily be backported to (streams).

When asked, Mike said (streams) isn’t going to go anywhere, (streams) is the stable one (it is stable again now), and Forte is very experimental.

March 12th, 2025. Just yesterday. This was the day that Forte saw its very first official release. And this was the first time that Mike talked about Forte in public as opposed to only to his immediate connections.

The family tree:

• Friendica (ex Friendika, ex Mistpark) (2010) (two new devs, relicensed to AGPLv3)
• Free-Friendika (2012-2012) (only fork not by Mike; created for there to always be an MIT-licensed Friendika; died because the sole dev couldn’t backport Friendica’s AGPL code into his MIT-licensed repository)
• Hubzilla (ex Red Matrix, ex Red) (2012/2015) (two new devs)
• Osada (2018-2019)
• Zap (2018-2022)
• Osada (2019-2019)
• Osada (2020-2022)
• Mistpark 2020 (2020-2022)
• Redmatrix 2020 (2020-2022)
• Roadhouse (2021-2022)
• (streams) (2021) (still developed by Mike)
• Forte (2021) (still developed by Mike)

Although he has solved account portability.