What is the latest experience with open app ecosystems?

I can start with some of my experience. One thing that was talked about early in this group was the need for vocabularies and protocols as glue for the smaller open apps, for flexible assembly. So way back in 2015 (can you believe it ?!) a group of us started on Valueflows, a vocabulary "for the distributed economic networks of the next economy, to coordinate the creation, distribution, and exchange of economic resources". It is nearing a stable v1.0.0, after much discussion and work and feedback from several projects that have used it.

In terms of the experience of creating smaller apps, we haven't had a lot of success. One partial success is that a few projects decided to use the same Valueflows graphql api between frontend and backend. This would theoretically allow fairly easy swapping of frontends onto different backends. Generally, there will be more frontends for many different user stories, while the backends are fairly generic. There has been coordination between groups on this. But.... nobody has felt a reason to try the swap yet. And a couple of those projects have started writing clients that skip the api and talk directly to the backend, partly because it is easier and partly for performance reasons. I think both of those projects plan on continuing support for the api, but it remains to be seen if this makes sense in the real world.

I could also mention a project that isn't particularly related to Valueflows itself, but has some interesting ideas on putting together a dashboard of open apps that can be related to each other. This is Moss, in the holochain ecosystem. It is a dashboard for apps from an agent-centric (person-centric) perspective, and supports groups and apps within groups. Perhaps most interesting is that you can connect pieces / objects with apps to pieces / objects within other apps in the dashboard. So for example in the case of a Valueflows app, you could set up a chat on a process that people are working on together, many possibilities.

Thanks for the update @Lynn Foster - I think it's vital to share both successes and failures. The dream of a network of interrelated software tools to support a new kind of way of being/living/economic relationship is a good one I think, but also perpetually seems to fail to produce fruit in much quantity.

For my side, working on https://karrot.world still (also birthed in 2015) - I have focused on making sure I am connected to groups with people who are using it for daily organising, which remains the case today. And slowly trying to move to a more "open app ecosystem" sort of concept perhaps. I added support for plugins this year (nobody has written one yet though).

A lot of my interest is not on the tech, but the groups themselves. I think the core struggle is not the tech, but how to actually organise with other people. And how to tell/share that story/dialogue/relationship. I think with a core of activity on the ground, the right tech can/will be created or appropriated.

So I got cautious with more utopian forms of tech that put a lot of focus on the potential of the tech. We just had a couple of weeks with the Karrot team in person exploration many wider topics, and one interesting thought experiment from it is that maybe Karrot isn't actually software at it's core, but a way of working and relating (within our team and with the groups and wider topics).

The tech solutionism, and tech bubble issues are very real I think. Most people do not want to relate to tech in the way I/we might, and sadly many projects which are explicitly about interconnection and collaboration are actually quite siloed and not collaborating very much! (not a blame, I understand how that unfolds, and struggle with that myself too).

A lot of potential still here though!

Thanks @Nick Sellen , thoughts worth exploring! I totally agree that your insight/focus on staying connected to groups using the software daily is really important.

Any other stories from other people that might be useful (or just nice) to share here? 🌱

Thanks for the opening the discussion @Lynn Foster and @Nick Sellen for jumping in with your insights.

One of the big pain points that came up in discussions here a few years ago was the sustainability of community-hosting organisations, to host instances of the servers that most open apps depend on. This bit me hard in 2020 when CoActivate bit the dust. They'd hosted a number of my projects for more than decade, including the Disintermedia blog and wiki, and their unexpected and permanent disappearance from the web was massively disruptive.

In the years before that, beginning with encouraging the Loomio team to incorporate as a Cooperative Company, I'd been getting increasingly excited about the potential of platform cooperatives as a potential solution to this problem. So in late 2022 as I started to get back to work after the lifequakes of 2020, I decided to gather some folks and set up a web hosting service as a worker-owned co-op called Bridge Seat, focused on hosting fediverse servers.

It's been a journey. We still don't have any servers up and running, so in one sense it's been a total failure. OTOH the only money we've spent is on a domain name, and I'm hopeful the relationships we've been built around the idea might still bear fruit. Over the last few months I've been in discussions with Andrew from Takahē about Bridge Seat taking over day-to-day development of his codebase, under his mentorship, and making it our first target for a commercial hosting service. Hoping to find funding for 3 salaries (lead dev, server achitect and me as community developer) to start work on this next year.

Otherwise, my main focus over the last decade or so has been raising awareness of open protocols/ standards (it's a bit out-of-date but I started mapping net protocols using MetaMaps). Bespoke APIs are great when you're prototyping things that have never been done before. But an open app ecosystem isn't viable if every app dev has to choose between dozens of competing APIs that do essentially the same thing, or implement a whole bunch of them, in order to interoperate widely. The ideal is to have a single standard for each thing that needs doing, which anyone can contribute to the shape of, and most developers agree to use.

I'm one of thousands of developers and activists who have been working on this for decades. Going back at least as far as BSD - which began as a standard distribution of Unix- and the GNU Project, which proposed that all software be written as generic standards, rather than proprietary products. Continuing through the internet standards defined at the IETF, and the web standards at W3C. As well as the informal incubation of standards like ValueFlows by folks like Lynn and @Bob Haugen, which is generally how experimental protocols are defined and developed, until they become stable and widely supported enough to become formal standards.

Advocacy for digital open standards involves both introducing more developers to the available standards, and evangelising the apps that implement them. This has started to really bear fruit for end users in the last few years. With anti-monopoly enforcement on both sides of the Atlantic (especially the EU DMA) pushing legacy platforms towards more use of open standards, and the massive growth of the Matrix, the fediverse and the ATmosphere, in response to the rapid enshittification of those legacy platforms. Getting to this point

But I'm not popping the champagne cork just yes. This has always been a 3 steps forward, 2 steps back, kind of dance. The DataFarmers will not give up their oligopoly control without a fight, especially now they've realised how politically powerful it can be.

Plus, even if governments confiscated all tech platforms from corporations tomorrow, and forced them to convert into platform co-ops owned by the people building or using them, there's still tremendous amount of work to be done to make a decentralised world of interoperating apps easy to navigate for the average person. That's one role I could see this group playing.