subgroups of subgroups

Hey @philippeponge can you describe a bit more about the case you have in mind?

Maybe subgroups of subgroups are the best solution, but it's hard to guess without hearing more about what you're trying to achieve.

it's hard to express with my English now details. Thank you @richarddbartlett

Fair enough, thanks Philippe!

Well I would love to hear from anyone that has thought "I wish I had subgroups of subgroups!"

I imagine there are a few different use-cases for that feature, so I'd like to have an idea of what different people have in mind, so we can come up with a design that works well for everyone

1. ORG MAIN GROUP 1.1 DEPARTMENT SUBGROUP 1.1.1 PROJECT 1 SUB-SUBGROUP

we wer just talking about this tonight as a matter of fact.

Cool @denjello :-)

I would be interested in seeing how we could organize subgroups in a more horizontal way so that we replace categorization and hierarchy with a structure that is more equal and in line with Loomio's principles. Just a thought experiment.

@philippeponge je pourrais traduire :)

@denjello can you describe to me what user experience you're imagining, e.g. what permissions do people have in each layer? I want us to think together about how people will use this, rather than jumping straight to an implementation.

@richarddbartlett I wouldn't be surprised if you're remembering the older discussion about parent group coordinator permissions. Potentially useful context for others here.

sure @richarddbartlett. Basically, the parent layer is public and allows for privat and public subgroups (and discussions). This solves the problem of a multi-faceted organization where EVERYONE belongs to the parent layer, and can apply to participate in child-layers. I envision a trickle up principle, where if you are a member of a sub-sub-sub group, you are automatically involved in every parent group. Is that clear?

@presley Ma demande concerne les sous-groupes donc.
Quand tu en as créé un, tu ne peux pas en créer dans celui-ci. C'est le principe de l'arborescence en informatique, bien pratique pour classer, ordonner :-)

Thanks a lot Presley

@philippeponge De rien! Est-ce que tu as besoin de cette structure pour résoudre quelque problème particulier? Je suis d'accord que ce serait plus élégant, mais je ne sais pas s'il est urgent.

I am interested in this for political reasons which can be seen here.

But even for our parties working page it would be great to split into sub sub groups. One of our members did this picture as an example.

@presley could you translate my precedent post ?
Maybe it's not easy ?

Bien sûr. Je pensais que ton premier message en anglais était assez clair que tout le monde a compris ce que tu as dit, mais je le traduirai.

"My question is about sub groups. When you have created one, you can't create more inside of it. It's the principle of trees in computer science, very practical for classifying and ordering things."

@philippeponge I have been reading your posts. There is a translate link in your posts, but it isn't active for @presley

If all else fails I use https://translate.google.com

No, @lbjoum - @presley wrote French while logged in with her profile set to English...

Thanks to Presley, Joum and Denjello for your translation and explications. I'll write directly in french without complex next time when it's too complicate for me in english.
Merci encore @presley car ta traduction me parait plus précise qu'avec Google traduction par exemple.

@lbjoum that's because @presley has her Loomio language set to English. It only shows "show translation" if your language is different from the poster's set language.

I have added this suggestion to the roadmap inbox: https://trello.com/c/YItPDGFp

I think that this is a good idea. It's come up a lot of times in my experience. However, I think it will probably be solved by answering some deeper design questions about how groups relate to groups generally.

Hello,
I haven't really been using Loomio, just watching for a while, so I haven't really grasped the philosophy behind it to any depth. I have a use case that I will tell anyway but I am not saying that multiple group levels is the right solution.
Here goes.
We are a non profit organisation for folk song tradition in the Nordics (Sweden, Norway, Finland, Denmark, Iceland) and the reason that we span several countries is that our languages in some cases are more or less different dialects of the same language.
Since we are different countries there would be national discussions below the common interest area. That would be one possible level. Sub groups would then be members/board, genres, professionals/historians and/or whatnot.
Below that there would be projects like Learning/Education/Courses and below that headline one would find the actual projects where we would diskuss and come to decissions.
As always one could flip these categories in any relevant order, but this example would show the principle.
In short: Organisation level / Geographical level / Subject level / Project level.
I am aware that this might be a much larges scope than you set out to solve so look at it ase a use case and not a proposition :-)

Best regards and thank you for doing a great job!
/chris

This is a complex subject and I don't think it would be fair for me (or anyone) to expect the Loomio core team to tackle it right now, as Loomio 1.0 requires lots of attention. With that in mind, I must say that once we start talking about subgroups of subgroups, and trees, we're getting seriously into nested hierarchies, and I don't think it's good for people to organize exclusively in that manner. I think that faceted classification with multiple taxonomies is more rational and ultimately more useful.

I would personally prefer for Loomio to look closely at (or to revisit) ideas such as tagged categories of people before it adds nested hierarchies such as subgroups of subgroups. Nested hierarchies do have a logical role in classification and design, but that role is limited, and often IMO grossly overused.

Tagging categories of people looks great! @gregorycassel, the discussion you linked to looks like it approved that idea, but you said it should be revisited. Does this mean that it's going to be implemented or not? Perhaps @alanna knows.

Yes I agree about tagging. Essentially we want to be able to easily group people into many, sometimes overlapping groups in order to meet organisational requirements, but without implicitly imposing a tree hierarchy.

It's something we'll be forced to address right after we meet feature parity with the new user interface.

I think taggiing is the best way to handle data on a technical level. All data should carry its own meta data and reside in a "great big pool". It is also a very good solution for search.
However treeview is a good and easily understood metaphor on the inteface level.
I am not saying that I think any of this should be implemented in Loomio. It depends on the scope of the project. A system that supports discussions and decisions with openness as a principle is a very good thing as it is.
But one could dream :-)

I don't understand the hierarchy concern. Just because you can nest doesn't mean you have to. We can already make lots of groups with overlapping memberships.

@presley the problem is that it's the dominant feature of group organisation right now and it really does not make much sense to be like that.

People look at it and think "Oh ok I'll come up with a hierarchy", just because the design invites you to do that.

@presley it's true that the memberships of groups, subgroups, and subgroups of subgroups could overlap in any imaginable way. The issue (to me at least) is that each discussion-- and any related 'Decisions'-- would only be in one group or subgroup. That's where 'nested hierarchies' or 'tree hierarchies' create separation (and at times alienation) which is probably familiar to all of us, but which is ultimately IMO troublesome and inefficient. The problems can increase exponentially with each level of hierarchy.

I think Alanna's OP in the old tagging categories thread I linked before did a great job of indicating some of the potential benefits of a tag-based system or subsystem. For instance, if you had tags including stuff such as @Wellington, @Dunedin, @remote, and @volunteers, you could create a discussion which would potentially be seen by all volunteers, or by all Wellington volunteers, or all remote volunteers, or all Dunedin volunteers. It just depends on which labels/tags are applied, and how they're connected or filtered.

By contrast, it's troublesome and inefficient to have separate subgroups for Wellington volunteers, Dunedin volunteers and remote volunteers. It's also troublesome and inefficient to have a separate group for volunteers, which contains discussions that have no geographic metadata (tags) whatsoever. 

With tag-based organization, one can avoid the practically impossible question of where to put a discussion which doesn't really fit into a given hierarchy. A discussion would simply be tagged with all relevant labels. Each person could follow whichever labels they choose to follow. Ideally, each person could filter and sort Loomio discussions according to any mix or match of labels they choose to follow regularly, or on a given day, during a given search.

Is this making sense? I'm rushing (sorry) and this is a complex subject.

I think everyone's excited about tagging. We haven't had any capacity to dig into designing or building it though.

There is an inherent difficulty in applying tagging in a Cartesian associative manner. You end up with a library catalogue of reference cards indicating the location of each instance of the tag. This can be very useful, especially for forensic discernment.

However, you lose all sense of 'flow'. The dynamic twisting divergence/convergence etymological connectiveness of dialogue and ideation. To capture this requires the added dimensions of graphical interactivity aspired to in 'Argument Mapping'

The ideal is to be able to use tagging to activate heat maps, stacked overlays, dendritic tapestries of threaded discourse etc. etc.

Part art, part UX, part UI, part fractile recursiveness. Strangely enough, some of us HF autistic types can do this in our heads. However, far more useful to have it available to share with everyone via digital medium.

Aspirational perhaps, but doable, even with current tech.

@Gray I think we've briefly discussed similar subjects before. My view is that etymological connection can't be precisely categorized or quantified. One argument doesn't necessarily lead to another argument, except insofar as individuals construct their abstractly reduced personal perspectives. Each person chooses and uses words in a different context on each occasion. This is especially true of words describing social concepts.

Having said that, I'm very interested in giving individuals a variety of tools to construct patterns, meanings and flow according to their personal perceptions. I'm definitely leery of any automated system, person and/or privileged groups who get to centrally formalize (in any way) the supposed shape of any topic's progression over time. I'm concerned about all forms of centralized authority.

Of course, any 'narrative' is also a subjective shaping of events. We do it almost all of the time. Typical forms of narrative, however-- including written stories-- do not implicitly suggest formal rigor and rightness in their shaping of events. People may write in rigidly authoritative styles, but that's not made inevitable by the toolset of writing in natural language. Nor need it be inevitable in visual representations; it just depends on how we create and arrange.

I really love maps and visual structure/organization. Maps of physical places can be precisely representational, but maps of ideas and dialogue should ideally IMO tend toward creatively decentralized impressionism instead of 'realism'.

"etymological connection can’t be precisely categorized or quantified"

Etymology can be mapped in a derivation sense. However, what is not so easily mapped are all the peripheral influences which help shape and contextualise dialogue.

Excellent British TV series by James Burke, 'Connections' detailed the often very disparate diaspora of ideation which connected the development of various technology of modern & ancient times.

One of the outcomes of endeavours like 'Argument Mapping' is to cater for ever evolving individualised pathways. Multiple ways of stacking, overlaying, creating specific inner transparencys etc within a body of knowledge.

Allowing for each individual to co-create their own maps of synergistic significance as per their needs. As with the art of infographics, they may then choose to share this visual insightfulness with others.

Thanks for the input, Greg

Thank you as well @Gray ! James Burke is awesome. This mini discussion has helped me to more clearly perceive my concerns about personally concentrated authority. I'm all in for personal creative tools and sharing tools. :)

Offtopic Divergence Warning

Suggest that issues are not so much re centralised authority per se. That is sometimes a logistical necessity.

It is the combination of lack of transparency around actions/decisions of said authority. AND Exclusiveness/Elitist behaviour around open access to authority.

Too often, the self-appointed few take it upon themselves to make self-serving decisions for the many.

Authority isn't 'evil', but the ways that it is misused! (great fan of 'servant leadership' model)

Offtopic continuation

I mostly agree @Gray : open and transparent processes are IMO the main medicines we need to treat the emergent ills of contemporary society. I do additionally strive for the radical minimilization of personally concentrated formal authority, within given spheres of (actual) stakeholders and investors. Like a number of people hereabouts, I favor collective governance, team-based project management, etcetera. However, balancing between collective governance and personal initiative (in 'executive' and creative actions) is one heck of a complex and contextual topic. Genuine servant leaders surely emerge in given groups, and for given goals. Regardless of formal roles or the lack thereof, people informally learn to personally and collectively invest time, labor and material resources in the initiatives of people whom they find reason to reasonably trust. This is a subject which some Loomio folks have taught me a great deal about.

If you (or anyone) want to discuss CParty stuff let's please do it elsewhere or privately greg.cass1@gmail.com . I'm not going that far off-topic here, haha. :)

It's difficult for me to read every things here, even with the traduction who is not really top actually.
I'll like to know if you are worked on this subject because I need it to order subgroups in one of the Loomio group I administrate.
To repeat my need, I would like to put in one subgroup two subgroups (or more) existing.
Thanks for your reading and future answering :-)

You are invite to correct my English in this message, thanks.

I don't think it is a priority for now @philippeponge but perhaps latter. I think the loomio developers like the idea of tagging more, which does create a similar ability to sub-subgroups.

It's really bad luck for our Loomio because I think it should be a good way for better ordered discussions like on a forum.

@philippeponge I know this isn't what you want, but it might be helpful for you to simulate an extra layer of subgroups, by creating subgroups which have names like these:
Group A: Subject 1
Group A: Subject 2
Group A: Subject 3
Group B: Subject 1
Group B: Subject 2

I've done it, look ! : https://www.loomio.org/g/ljU2gKpH/democratiereelle-ile-de-france

What I need is now :
Group A : - Subgroup 1 (discussion 1 / discussion 2 / ...) - subgroup 2 (discussion 1 / discussion 2 / ...) - subgroup 3 (discussion 1 / discussion 2 / ...)
Group B : ....
Group C : ...

Thanks a lot for your interest and your trying to answer me :-)

The group looks very interesting @philippeponge -- my kind of subjects. It really makes me wish I knew French!

@philippeponge I wish we could build this for you :) Please find us some more software developers to help.

Opportunity to motivate research, well done @alanna ;-)
What language do you develop ?