Three Months with Loomio

"We want to solve tomorrow's problems with today's tools instead of treating today's problems with yesterday's tools."*

---

Introduction
In this discussion I desire to maintain a space to collect the steps we will be using to empower our community with Loomio. I will remain honest about our situation (without compromising privacy concerns) and my approach toward group adoption. My hope is that through this documentation it will be possible for our team to reflect on the process and in an extended sense offer a glimpse into methods that collectives can use to refine their decision-making process.

Backstory
Our non-profit organisation (Gängeviertel Hamburg) made a consensual decision in December 2014 to try Loomio for three months. During the discussion leading up to consensus, we were reminded of the failed experiment with liquid democracy, email accounts overflowing with messages from over 30 different mailing lists, the challenge of reaching those of us who never touch a keyboard and perhaps most importantly, the privacy concerns that all cultural activists in the 21st century are wise to share.

The Way Forward
Nevertheless we decided to give the green light to the project for a test phase, but on the condition that it remained an internal tool for resistance minimisation and centralised discourse management. This meant that only active members of our organisation would be invited to participate, and furthermore necessitated a private installation of the Loomio codebase with on-site storage.

WTF is Loomio
I'll send everyone a nice mail about registering. Several of the topics at hand will be discussed. We will set the voting deadline to 60 minutes prior to our weekly face to face meeting. At the meeting we'll use the discussion on Loomio to inform the decision making process, but will be careful not to interpret Loomio as the decision.

First things First
After assessing Loomio for the first time, I knew that I was going to have to get jiggy with the code here and there. On my to do list was castrating the AWS image upload system, introducing some crossbrowser CSS fixes, drop in my live markdown editing suite, rebranding (including color, logo, icons and footer links), as well as tossing in a few routes to our own home grown node-js Archive. But first I'd have to get my feet wet and fork the codebase. And spin up a VM. And clone the fork. (Thankfully the howto at github worked flawlessly on the Ubuntu 14 VM). And install Rubymine and reclone the git repo on my dev machine. And rebase my ruby installation. (How many versions of Postgres are running on my laptop? 3,4???) And then .git_ignore my modded Gemfile cuz my devbox is debian and there is only 2.2.0-dev for testing.) And then get the smtp server working for outbound. And then sleep.

Along the way I patched a recurring issue (the avatar duplication thingy on the profile page), caught a transcription error in the global locale string "gravtar" and completed the translation of the German locale (it was only two strings away). I solved the problem of uploading avatars and banners to the filesystem (instead of AWS) and fixed some weird recursion errors that forced me to scroll up the terminal to figure out what was happening while things were borking. Did I mention that I have never really used Ruby because I am actually from the javascript camp, that I prefer jade to haml and think less beats scss any day of the week? At least Angular is on the horizon (and this forces me to finally overcome my most-likely juvenile abhorrence of coffeescript).

And then I decided to read all of the open Issues on GitHub. I commented on each and every one of them, linked a couple duplicates to each other and offered some solutions to a couple. Maybe it was a bit of overkill and I hope I didn't make any enemies, but at least this way I'll be informed as soon as the issues get updated. Usually when I consider forking and developing a project and see that there are that many issues it usually makes me run the other way. Luckily I already "committed" to participating in Loomio and although it took forever to read all of the issues and think about them, it was a really good experience because now I think that I have a pretty good grasp of the pitfalls ahead and admiration for the great work done by the developers to date.

Of course there were concrete effects too: Several open issues were simply closed because they were old (and the issue reporter just needed a reminder to close their issue) and I tracked down a logic fallacy in constructing the users preferred language when visiting the website. This led me to a better understanding of the code, and is something I can totally recommend when jumping into any open-source project. :)

Here is a list of most of the mods I have enabled on our system:
1. @username mentions in comments link to /u/@/username which displays the @username profile page as mentioned here
2. On the profile page the email address of the user is displayed (never mentioned anywhere, just did it.)
3. The avatar thumbnail fix that I already issued about here
4. Markdown Buttons to help non-coder users as talked about here
5. Language Selection Logic Fix as talked about here
6. Search Display Fix as implemented here
7. Turned off AWS for preference of localstorage as described here
8. Turned off Attachments to Comments because TOTALLY B0RK3D.
9. Turned off Facebook, Twitter, Google login because I COULDN'T BE BOTHERED

A Passenger with lots of RAM
Well, switching to production from development could have profited from a "testing" phase - but since this whole thing is a testing phase we decided to just jump in - feet first. The first thing that was clear was that the dozens of configurations sprinkled here and there didn't make it any easier. I rerouted / to /users/sign_in because the landing page didn't hook into the css and js assets like everything else and I was out of time. Definitely room for improvement - but also a symbol of what you can expect from a system that has been in iterative development for over 2 years... Maybe its time to clean up the distro and package it as a complete docker app that interactively asks you for things like your hostname, secret_key etc??? At any rate, we're live and I survived. So that's something. Let's just see how the users come to grips with it...

Adoption
So I got the mail system to work and then I sent out an invite to 60 people and the server crashed (no clue as to why). Then I got a few emails from people from outside of the intranet that they were only getting some weird message from Apache, but after fixing some routing stuff in /etc/apache2/sites-enabled/ and the reverse-proxy box we were finally online. At the moment I am creating individual groups for each "department" and inviting people personally. I did already get some feedback like:

"Why am I not automatically subscribed to every discussion in the group that I created?"
"Why do I have to turn on Markdown?"
"Why do I have to know every persons email to invite them to a group and what if I use the wrong email address?"
"How can Invite a bunch of people?"
"Can I vote / comment by SMS (text message)?"
"Why is it so complicated?"

Waiting for the first Proposal
After getting some people into the system, I noticed that there was a lack of interaction in the public groups that had been created. It turns out that on the /explore page, only groups with three discussions are listed - which was a strange edge case and one that got me back to thinking about the chicken and egg problem that I had seen elsewhere and we discussed internally thusly: "At the beginning, our loomio is like a party that is empty. People come and see that no one is there, and then leave because no one is there." So I turned off the feature that rate limits groups this way and our participation skyrocketed within a day.

The first Proposal
So we finally have our first real proposal on a topic that is not really that important (it is a decision about a suggestion made by an architect for the redesign of the entrance of one of our buildings). At our weekly meeting last week, there was no consensual decision to be made because it seemed like everyone was split between the two options, which I'll call the "old" and the "new". So I extended the deadline for the decisions of the proposal for a week and invited everyone to participate. Because we take preservation of historical buildings seriously, I knew that this would get some attention. Indeed, 32 comments (including decision backgrounding) qualifies in my perspective as participation. Several people have even changed their vote along the way as a reflection of the friendly discussion.

Which language do I use?
On a sidenote, I speak several languages but prefer English for interfaces and day-to-day writing. That said, most of the members of the organization speak German as their native tongue, some speak Spanish and others French. I try to tell myself that I will only write English when my profile is set to English, but sometimes I forget and write in German...

USAGE STATS

---

Weeks ago	New users	New top-level groups	New discussions	New comments	New proposals	New votes
0	15	1	8	54	2	29
1	6	1	2	6	0	0
2	20	3	5	17	0	2
3	4	5	2	1	2	4

---

* You should really read about Parkinson's Law here first and then here

* The quote at the top is modified from a talk by Pia Mancini