Expanded 'trinity'

This is to discuss the thread 3 Expanded ‘trinity’ - A generic toolstack for distributed organising and knowledge collaboration emerging from the meeting on Shared learning spaces 15mar2021. The emerging threads are summarised here. Possible participants include: @asimong @mikemh @wouter. Maybe @robert.best Maybe @Jonathan? Maybe @osb (meet.coop, OpenCoop), @mnoyes ?

‘Trinity’ refers to a framework developed by Rich Bartlett & Nati Lombardo and adopted by Guerilla Translations as part of the DisCO ‘stack’ of tools. A year ago this was explored as a ‘toolsframe’ at Open2020 in this slide deck and speaker notes. The trinity/stack model is referred to in passing in this slideshow from the 15march discussion (see slide #4) and this recorded presentation. But as suggested in the 15march discussion, the model of the ‘trinity’ and the larger toolstack that it belongs to now needs to be updated and defined in more detail - work to be approached in this thread.

Happy to chime in here. What do you think, shall we start with agenda-setting, or start with introductions / understanding where we all come from, how we are approaching this? Would be good to hear from everyone who would like to participate, check-in style, before getting into the detail of the conversation.

1 Like

Thanks, Mike, just that I’m not so sure about the efficacy of doing these in fedwiki space. I would prefer a more directly collaborative approach e.g. HedgeDoc.

I’m not so sure about the efficacy of doing these in fedwiki space. I would prefer a more directly collaborative approach e.g. HedgeDoc.

HedgeDoc and fedwiki are rather different in what they enable - HedgeDoc being just Markdown docs (which I like, and use), and fedwiki being primarily a means of sharing systemic ‘maps’ of stuff; also discrete chunks of content (including chunks of words). Once you know how to use it, fedwiki is intrinsically more conversational - actively federating - than HedgeDoc - which is just a shared document editing space, yes? The catch is - ‘once you know how to use it’ :roll_eyes: There is much more than meets the eye of a reader, but becoming a fluent writer-collaborator takes some hands-on time.

I’ll not make a move in fedwiki just yet (current time limitations) and haven’t opened a thread on this yet in the forum. But I’m certain that this richer, more ‘mutualist’ medium enables collaboration in ways that more familiar, passively ‘shared’ media don’t reach. Watch this space? Maybe in the coming month.

We’re on ‘Expanded trinity’ ground here, so I’ll move this exchange to that thread.If I can manage that feat of curating :wink:

To be sure, the affordances are different. What I am suggesting is careful considerations of the match between affordance and purpose. My view at present is that the affordances of fedwiki are very specific, excellent for some particular things, but no good for building collaborative documentation. And I appreciate the eye-roll!

Yes, I think fedwiki isn’t primarily about documentation - like orthodox wiki, for example, is. I’d say fedwiki is primarily about collaborative thinking, exploratory mapping, trying-out understanding - done between known collaborators, not for an undefined ‘public’.

My sense of what we might be wanting to do in the Community programme, is that it includes a lot of trying-out and mapping (‘r&d’), as well as an amount of documenting (of known resources, etc). But as I wrote earlier, I don’t want to engage this fedwiki thing abstractly on the text page, it needs hands-on, workshopping. So right now, practicality means I’m wanting to hold back on this particular tool question.

It’ll be good to see folks chiming in with your earlier broad query about approach.

I agree. The header to this thread contains some background, on the model referred to as ‘the trinity’. But let’s check in with individual perspectives on ‘tools’ and architecture of tools, before we get into models and modifications of models? I do believe trinity etc is a good starting place, needs some evolving though.

Hi @wouter @mnoyes @asimong - Is it time to kick off this sleeping thread? Others: @Jonathan @osb @jamie @freescholar ?

I made a start on sketching ‘toolstack’ thinking in a fedwiki, here Digital toolstack. And would be glad to have an ‘enquiry pod’ start up, to explore using wiki as an investigativemedium for long-form thinking, alongside this forum for commentary and chat. @robert.best may have a hunch about this kind of ‘pod’ investigation process?

You probably spotted . . that’s a working version of toolstack-as-learning-space :wink: ie Repo (wiki) + Threads (Forum). Do we need the third trinity element too - Chat - or is that adequately covered here in Discourse? I offer my Zulip ‘conviv’ space for chat, if needed: conviv Or we could drop into social.coop Mastodon https://social.coop/

‘Extended trinity’ is a whole lot evolved from the ‘trinity+stack’ model described in 2019 by Stacco @ DisCO Handbook - more nuanced, more 2021 up-to-date? But I feel they have the same basic rationale, and it’s a good one. What do ppl think? Is there another frame out there, for reviewing and configuring a collaboration toolkit?

Some folks find the whole ‘browser is your toolbox’ approach unhelpful, fragmentary? Maybe protocols is where the action is, rather than the apps that float on the protocols? Both-And, imho. If that becomes an issue, we might need to fork this thread . . But: let’s get started?

@asimong wants to encourage check-in, as well as (premature?) deep-diving. the stuff above will serve as my check-in :wink: As others chip in, check-in too?

I’m preparing a session in the Social Economy Master programme I’ll be doing tomorrow, on “Methodologies for co-production and ICT tools for the Social and Solidarity Economy”, at the Barcelona Autonomous University.

I’ve added in your slide 4, @mikemh to discuss the extended trinity. My slides are in catalan, but basically have this ToC:

  • the Commons
  • recentralisation of the Internet
  • Extractivist platform economic model
    • GAFAM
    • Networks: descentralisation and recentralisation
    • Responses
  • New ways of organising
  • The Five Pillar model of sustainability
  • Don’t be a bottleneck!
  • Stigmergy
  • Case: Som Energia
  • Tools and methodologies of managing projects
  • Communication and management tools for the commons + SSE
  • Trinity of tools (x3!)
  • Case: CommonsCloud: a cooperative cloud
  • Case: Meet.coop
  • What do we take with us - FEEDBACK ROUND

yesterday I had this session with 20 adult learners in the master programme on Social and Solidarity Economy of the Universitat Autonoma de Barcelona. It’s interesting because it is coordinated by a local network of SSE members; therefore also the learners are mostly members of social economy orgs and initiatives. Here’s some feedback.

That said, it’s just a 2 hour session I do once a year in their programme, so no time to really get into much detail or handson work with them. Fortunately the commons part was already treated by a colleague of femProcomuns in a previous session, so I could take it from there.

I did weave the trinity and extended trinity into the session and discussed a series of tools that could be used to resolve parts of the trinity.

Participants of the session were vary positive about the session; main questions that returned several times: 1) do you need to be very techie to be able to use these tools?, 2) will the rest of my team want to make the change?

I encourage people to try it out as much as they can. Here it is really helpful to have cooperative clouds and other ethical services that provide webapps as a service. Most collectives lack an IT department or even techie team members, so that’s key. The second is really a human habit against change, where teams work in ways that made sense a while ago, but cannot cope with new people joining the process very well. Stigmergy would help the new people find out what had has been decided and why, what tasks are waiting to be taken up so they can make a contribution without wasting too much coordination efforts with the team. A culture of change and frequent evaluation and reflection is needed, but not everyone is ready for it. Even changing from Whatsapp to Telegram to Element.io is something that people at first are afraid of. Once they have taken the step, it is often not so bad or even a liberation.


One way to have an inexpensive cloud account with tools built in (Tasks, Notes, Bookmarks, Forms, Calendar, Circles, Photos, Markdown docs, OnlyOffice docs/spreadsheets/presentations) is to join May First Movement Technology. NextCloud is part of the basic tooling. The following details are from here: May First Movement Technology member benefits

A $50 US dollar annual basic membership applies to all members. This basic membership fee is the minimum contribution necessary to join our cooperative and include basic service benefits of 1 e-mail and 1 Nextcloud account with 10GB of storage between them.

If everybody in your organisation/group were in that share, it could make so much difference?

Other coops provide equivalent services of course - obviously, WebArchitects, Collective.tools, CommonsCloud - but it’s not simple to run comparisons on facilties and fees.

MayFirst seems to have a reasonable offer at least for individual members. As a collective maybe a dedicated NextCloud instance would be more desirable?

There’s also the aspect of Local Economy & Language. Most people on earth are in need of communities that speak their mother (or father’s) tongue.

In Barcelona we have a few ethical and social economy cloud providers:

  • CommonsCloud.coop (coordinated by femProcomuns.coop): provides Discourse, NextCloud, Phabricator, LimeSurvey, Zimbra webmail
    • individual members pay 4€/month for 10GB in NextCloud
    • collectives pay 30€/month for their own NextCloud instance with 100 GB included
  • Pangea.org, an association similar to MayFirst, providing email accounts, NextCloud access, …
  • Maadix, providing a server that allows collectives to manage their NextCloud, Mailtrain, WordPress, mail servers, mailman, etc through a grahical webinterface. It is privacy-by-design and the collective manager has root, without having to enter through the commandline. The founding woman runs it as a Ltd company, but is fully committed to free culture for long.

Possibly we could make a table with ethical webapp service providers and accessible ways to use or simply test out each tool. If so, we’d need to include the region/language for sure.

:heart_decoration: I’d like to follow the principle of this. A couple of ‘methodological’ challenges then spring to mind:

  • ‘A table’. I think that the necessary comparisons are more multi-dimensioned than a table can properly handle, for visualisation purposes. So this raises the question of tools for mapping/visialising (extended trinity). Something ‘graphy’ and interactive, overlaid on a basic database (which, yes, could be a spreadsheet). Like (proprietary :frowning: ) Kumu? Christina Bowen @ DigLife is good on Kumu, interactive graphical resource mapping.
  • An exploratory, prototyping, mapping exercise, to discover an analytical frame for selecting of toolstack-hosting platform services. Based on what sample of service providers? To include variation in Local Economy & Language . . How about sampling by taking the services locally used by and known to half a dozen members of meet.coop? Barcelona @wouter , Brighton @mikemh , Belgium @asimong , Sweden @petter , Toronto @Yurko @robert.best , NY @jamie , Boston @freescholar . Plus at least two south of the equator @fredsultan Jason Nardi (not here in the forum yet?) @camille ?

MayFirst will host for a collective (as will WebArchitects, and many other free software host coops). Many members of MayFirst are collectives (ie movement organisations) and being geeky, some of these use MayFirst’s custom hosting offer. So the mapping would need to include the dedicated/custom hosting services that are available, as well as ‘off the peg’ toolstack platforms. Two very different offers?

Personally, in searching for my own toolstack providers these past couple of years, I find the offers of custom hosting utterly mystifying. Not being a sysadmin, I need a service with no sysadmin work falling to me. If I want a wiki space, for example (extended trinity, ‘repo’), it seems that I must self-host it. No way. As distinct from Zulip, say (extended trinity, ‘issue threads’), where I can spin up a space on demand. If I want Discourse (again, issue threads) I have to self-host it. Whereas if I want Loomio (‘issue threads’) I just need to pay a fee - which for a voluntary organisation with no funding, can be done for a lifetime low fee.

The issues of price-per-service are really hard to get the head around?

My impression is, most free software hosting offers are too geeky. If I can use the analogy of a finding a place to live, they seem to be offers for ‘rental with furniture’, not ‘fully serviced’ living arrangements. But I admit, I’m confused by exactly what the offers are. The websites seem to be written by sysadmins, for sysadmin clients.

I would say, removing this element of fog - thro clear descriptions, but better, thro clear hosting offers of ‘full service, no sysadmin burden’ - could make a lot of difference to the deployment by ‘ordinary’ organisations. Does eg Cloudron achieve this? I’m not sure. Am I dim, or are the platform-service descriptions inadequate (including, for example, Cloudron), and hostage to geekspeak. It can’t be just me, after a couple of years trying to get to grips with this toolstack platform-service challenge?

Downside: ‘full-service’ would not be ‘cheap as chips’ - which many users might naively hope for? The labour of curating the tools and running the platform must be paid for - as meet.coop ops members know only too well. So: the perception of free software = free beer also needs to be swept aside? The contribution model is critical and, in the short term, might involve higher service charges? I can’t believe how cheap MayFirst’s offer is :roll_eyes: (Although . . this has recently been recalculated. So let’s assume it’s well judged?) Whatever . . the expectations of toolstack-platform user-clients need to be evolved, to match the economics of contribution (aka sweat equity)?

Agreed that many free-as-in-freedom services are offered in a somewhat or too geeky fashion. It is certainly a challenge to build these services as a sustainable - thus long term - business, as we know from each of the mentioned initiatives and meet.coop itself.

About Cloudron: in 2017 when we were doing R&D for CommonsCloud we tested both Sandstorm.io and Cloudron.io Each with their advantages, and also each with some drawbacks. We felt we didn’t have full control over their user experience, so decided to simplify. In any case, Cloudron expects you to have at least one (junior) sysadmin person, or at least someone who can configure the services through the webinterface. That’s not bad for some collectives. IIRC DigLife runs a Cloudron instance for their collective, isn’t it?

  • MayFirst: 25US$/yr for individuals and 50 US$/yr for orgs, but orgs have Dues added depending on their annual revenue
    • a 50 US$/yr subscription entitles you to 10 GB (tier 0)
    • Tier 1: $125.00/yr → 60G
    • Tier 2: $250.00/yr ->150G
    • doubts: can orgs manage their own users in a NextCloud instance? (be it shared or dedicated)
    • at CommonsCloud they can manage their own collective, add up to 150 users. The minimum for collectives is 390€/year with 100GB, so more costly than MayFirst
  • ProtonMail: 4€/month for mail incl 5 GB + now with fileshare included
  • About ethical mail services we created this comparison table a few years ago
  • When you buy a mobile from the e.foundation you get it with their customised version of LineageOS (with great improvements) and an account on their email + nextcloud servers

Actually for our colleagues in the South, to find a secured server to host Nextcloud is a big issue. My colleague in Latin America are setting up their own Nextcloud with a server based in Europe. I have a question actually about the location of ther server used for BBB by meet.coop. Is the server in Europe or are there several servers ? We are thiking on how we can support each region with a server closer to get better connection for Nextcloud and also for BBB.

Your regional strategy with servers sounds a good one @camille .

Having regional servers is an ambition for meet.coop, I’m sure. But technically, I imagine supporting this kind of expansion in the infrastructure would be difficult at the present stage - tech labour is very stretched in the coop. A 99vanAmsterdam funded project is supporting a current burst of development and learning, which will be very useful to the coop, downstream. But this kind of expansion would probably not be possible right now without that specific funding.

The meet.coop public BBB server is in Canada. There is another in Germany which is used in a more limited way, but afaik the coop does not yet have experience in deploying the technology for load sharing or routing across servers. ¿@wouter @dvdjaco know the situation better, including possible links with free software hosting coop/s in South America?

The matter of multiple regional BBB servers has gone quiet in recent months. People are busy on other things? @freescholar has an interest in this too, in North America. Are regional servers an issue in the North-South Francophone networks @fredsultan ?

https://mayfirst.org hosts Nextcloud and we are discussing hosting BBB, right now only jitsi is hosted for video chat. MayFirst servers are in NYC and Mexico City.
Agaric hosts our BBB server in Maine, bu has access to servers around the world with our current provider. We do not have the funds to host our own BBB server right now, so we are not yet part of the hosting with meet.coop but working on it with MayFirst possibly expanding to offer BBB in the future.

As @mikemh says with meet.coop we started with the first BBB server in Montreal/Canada (ca.meet.coop), with DigLife we run the demo server in New York (currently there’s some maintenance issues with that one), and some months ago we launched the first server in Europe, in Germany (de.meet.coop). In that same datacentre we also set up a cluster for the City of Amsterdam for the Cities for Change. There are several improvements that we are implementing in that last server (cluster). And the idea is that later we can deploy some of these improvements to the shared servers.

For operational efficiency it would be easiest to have one cluster of servers in one location, but that’s not good for the user experience. The thing with online meetings is that for decent audio quality there shouldn’t be too much delay. Therefore the server location wrt the participants location is important. So we end up needing regional servers, and when we grow: regional clusters of servers.

So to say meet.coop’s regional strategy is to grow organically: when we have enough members in a certain region/continent and appropriate local partner we’ll want to set up a new server or cluster.

In Latin America we have a series of contacts of free software collectives that could be of help. But we can only explore (seriously) this if we start to have some users and contributions. In Australia & New Zealand as well and in India possibly. In Africa I’m not sure; when the needs arise we’ll have to digg up old contacts and explore collectively.

Another challenge is enabling members to freely choose between regional clusters. Once we have Single Sign On, as we put it in the roadmap, then we can allow every member to choose and/or use all clusters as they wish. Note that loadbalancing is on a cluster level. It’ll be the members who can choose where they want to host their conference rooms (in what regional server or cluster of servers). When we get there we should do some work to make these choices straightforward and intuitive, but that’s another issue. That will mean a serious improvement for members and very much for membership management as well (that’s something we do manually right now).

1 Like