chat
Original discussion starter:
Have u discussed adding group and subgroup chat as new way of communication beside discussion?
Often brainstorming and warmup through helps to formulate the discussion ..
Proposal for community-led research and evaluation
Existing free code chat software could be integrated into Loomio as a module, or a Loomio server could be connected to a stand-alone chat server through an API, with the chat integrated into the Loomio UI. Some chat code/ protocols are designed with one-to-one chat in mind, with conferencing added on as an afterthought (eg SIP, WebRTC), while others are designed around group conversation in a "room" that can persist with or without having active users in it (eg IRC, Mumble). The latter type are more suited to the Loomio use case, and more likely to scale up to the numbers of users active on a Loomio server at any one time.
Chat modules (or apps containing them) for evaluation
- Name | Demo | Source Code | License | Language(s)/ Framework(s) | Standard(s)/ Protocol(s) used | Web-based/ Server-client/ P2P
- Etherpad | Demo | [Source Apache 2.0 | JavaScript, Node.js | ? | web-based
- GNU Ring | ? | GPLv3+ | Source | SIP/DHT | C, C++, Java | IP | P2P
- Jitsi Meet/ VideoBridge | Demo | Source | Apache 2.0 | Java, Javascript | WebRTC | web-based
- MatterMost | Demo? | Source | AGPLv3 (server), Apache 2.0 (admin tools) | Go | ? | Web-based
- MetaMaps | Invite-only Beta | Source | AGPLv3 | Ruby on Rails | ? | web-based
- Mumble | ? | Source | BSD 3-clause | C++ | VoIP, Mumble protocol | Server-client
- Palava.tv | Demo | Source | LGPLv3+ | Ruby, Java, Javascript, JSON, coffeescript | WebRTC | Web-based
- Rocketchat | Demo | Source | MIT | Javascript | Meteor/ SAML, Jitsi VideoBridge | Web-based
- Riot.im | Demo | Source | Apache 2.0 | Java, Javascript, Objective-c | Matrix | web-based
- Tox | ? | Source| cross-platform [Clients] in a range of languages; C++, C, Vala, Python, Scala,Objective-c | Tox protocol, DHT | P2P
Pierre-Yves Mon 3 Nov 2014 6:36AM
In my humble opinion, the chat should not be integrated in loomio (at least at this stage; there are many other improvements more important than this). The groups can use a 3rd pary (video)chat if they need so.
Matthew Bartlett Mon 3 Nov 2014 7:15AM
Here's one of the times this was discussed previously: https://www.loomio.org/d/UHAqZM01/private-message-system
Wael Al-Saad Mon 3 Nov 2014 7:28AM
Text Chat like the one at etherpad
Wael Al-Saad Mon 3 Nov 2014 8:07AM
I see .. the idea behind that, tools commons empower the crowd
Balasankar C Thu 6 Nov 2014 9:29AM
I suggest we use XMPP for chat so users can use their existing accounts.
Danyl Strype Tue 30 May 2017 4:19AM
I notice that MataMaps have recently added a live chat sidebar with a similarly user experience to the chat sidebar in Etherpad. It can be expanded, or hidden leaving only a button on the side of the screen. Not sure what protocols the MetaMaps chat feature uses (Websockets? WebRTC? Something else?) but it supports voice and video.
AFAIK MetaMaps is fully free code, and MetaMaps uses Ruby on Rails like Loomio does. Perhaps their chat systems could be modularized so it can be used by other free code projects like Loomio to add a chat sidebar?
EDIT: Removed reference to Hylo. What they've added is a direct message system, not a live chat system, and it isn't working for me.
Z. Blace Tue 30 May 2017 6:57AM
interesting. agree live chat is much needed feature. if loomio wants to be standalone product it is even urgent.
James Kiesel Tue 30 May 2017 11:32AM
I'm fairly sure we won't be working on a chat feature in the near future. I've talked in a couple other places about limiting Loomio's footprint to deliver a really great experience at our core competency, which is decision-making. Loomio was never going to rival Facebook, or Discourse or Basecamp or some of these EAS tools which just do freaking everything under the sun (although we thought it could when we started out)... we do not have the resources to support such an endeavor.
Here's a list of chat apps I can think of off the top of my head: KakaoTalk, Line, Messenger, Slack, Signal, Discord, WhatsApp, RocketChat, Skype, Voxer, Yahoo messenger... I'm probably missing a dozen more.. you probably have at least 3 installed on your phone right now. Users are exceptionally spoiled by their (usually excellent) experiences on these apps, and it's basically just not an area I'm interested in competing in at all. I've actually built a chat module for Discourse a while back; it's a non-trivial endeavor to get right, and it would be really easy to sink a whole bunch of time into a mediocre-to-decent chat that nobody ends up using because there are so many ways to communicate on the internet.
So, I think it's a much better use of our time to position Loomio as a slide-in, lightweight decision-making tool which can be used easily from these other tools... things like 'drop a deep link to a poll into a chat anywhere', or 'make a decision with your slack team without ever leaving the Slack app', rather than 'support short informal discussions in a tool that's designed for relatively formal communication'.. this is the thing that no one else is offering, the 'missing piece of the internet' that we talked about in our crowdfunding videos and elsewhere.
I'm, of course, really open to having my opinions here challenged :)
Danyl Strype Wed 31 May 2017 12:47PM
This bit I want to challenge is this:
"it would be really easy to sink a whole bunch of time into a mediocre-to-decent chat that nobody ends up using because there are so many ways to communicate on the internet."
This assumes that Loomio would build a brand new chat system from scratch, and that it would therefore be a mammoth effort. But you don't need to roll your own (and maintain it). Free code for web-based chat systems already exists, as I pointed out with the example of MetaMaps. You could build on this work by making a chat module a dependency of the Loomio codebase itself, or creating a chat API that allows the admins of each Loomio server to hook it into the chat back-end of their choice.
That would require you to come up with some evaluation criteria, evaluate what exists (Kalashnikov principle), make a choice, and figure out how to integrate it neatly and intuitively into your UX. Loomio core devs don't need to do all of this work either. It seems like there are at least a few of us in the wider Loomio community who want this. If we can work up a list of criteria here, or at least the information needed about each candidate (language/ framework/ protocols/ license etc), the community can help with testing stuff (from user and/or admin POV) and adding all the info for each candidate to a list in the description box. I'll start with MetaMaps to demonstrate what I mean.
KaNNoN · Mon 3 Nov 2014 6:09AM
videochat?