Mattermost's future: who wants to fund/develop a fork?

EN: Mattermost’s future: who wants to fund/develop a fork?

TL;DR: Following Mattermost’s restrictions, we created a fork (Mostlymatter) but we lack resources. We want to connect funders and developers to maintain a free and robust fork.

Hello,

As you may (or may not) already know, we are concerned about the future of the Mattermost software (an alternative to Slack, among others, and the engine behind the Framateam service).

At Framasoft, we love Mattermost. It’s a great piece of free software: fast, efficient, etc.

To our knowledge, we host one of the largest Mattermost instances in the world: Framateam (180,000 users, 34,000 teams, 200,000 discussion channels, nearly 60 million messages, etc.).

This software is part of the technical and political commitment that Framasoft has been promoting for over 10 years: de-Googling the internet (and, more broadly, fighting against surveillance capitalism and campaigning for an emancipatory digital world).

However, for several months now, the Mattermost user community has been concerned: last October, Mattermost announced significant restrictions on these community models: https://forum.mattermost.com/t/mattermost-v11-changes-in-free-offerings/25126

This prompted us to create our own fork, Mostlymatter, which was publicly announced and recognized by the Mattermost team.

PLEASE NOTE: this message is NOT intended to criticize the commercial orientation of the company that publishes the Mattermost software. A lot has already been said on the subject (here or there, for example). It is the company’s choice, and ranting about it here will not help anyone (we will moderate any aggressive/derogatory comments about the Mattermost company).

As the authors of Mostlymatter, we have been contacted by various people asking us to add different features (such as SSO) to our fork.

We have made it clear that we did not have the energy, the skills in Go (the language in which Mattermost is written), or (above all) the time necessary to accept external contributions, or even the services that were offered to us.

We are writing this message so that we can freely discuss here among individuals/organizations/institutions who do not want to find themselves in a community deadlock.

In particular, we would like to facilitate contact between:

  • individuals/organizations/institutions that would be willing to finance a “solid” fork of Mattermost/Mostlymatter (note, however, that Framasoft is willing to contribute a small amount of money, even though we are a small non-profit association with limited resources)
  • individuals/organizations/institutions that would be willing to maintain and develop such a fork (paid services, therefore)

We are posting this message in French and English but encourage responses in English.




FR : Avenir de Mattermost : qui veut financer/développer un fork ?

TL;DR; : Suite aux restrictions de Mattermost, nous avons créé un fork (Mostlymatter) mais manquons de ressources. Nous voulons mettre en relation financeurs et développeurs pour maintenir un fork libre et solide.

Bonjour,

Vous le savez peut-être déjà (ou pas), mais l’avenir du logiciel Mattermost (alternative notamment à Slack, et qui motorise le service Framateam) nous inquiète.

À Framasoft, nous aimons Mattermost. C’est un logiciel libre formidable, rapide, efficace, etc.

Nous hébergeons, à notre connaissance, l’une des plus grosses instances Mattermost au monde : Framateam (180 000 utilisateur⋅ices, 34 000 équipes, 200 000 canaux de discussion, près de 60 millions de messages…)

Ce logiciel fait partie de la proposition technique et politique portée depuis plus de 10 ans par Framasoft : dégoogliser internet (et, plus largement, lutter contre le capitalisme de surveillance, et militer pour un numérique émancipateur).

Or, depuis plusieurs mois, la communauté d’utilisateurs de Mattermost est inquiète : en octobre dernier, Mattermost a annoncé des restrictions fortes sur ces modèles communautaires : https://forum.mattermost.com/t/mattermost-v11-changes-in-free-offerings/25126

Cela nous a poussé à créer notre propre fork, Mostlymatter https://framagit.org/framasoft/framateam/mostlymatter, publiquement annoncé et reconnu par l’équipe Mattermost.

ATTENTION : ce message n’est PAS destiné à critiquer l’orientation commerciale de l’entreprise éditrice du logiciel Mattermost. Il y a déjà beaucoup, beaucoup d’encre numérique qui a coulé sur le sujet (ici ou , par exemple). C’est le choix de l’entreprise, et rager ici n’aidera personne en rien (nous modérerons tout propos agressif/dénigrant envers l’entreprise Mattermost).

En tant qu’auteurs de Mostlymatter, nous avons été contacté par différentes personnes qui nous demandaient d’ajouter à notre fork différentes fonctionnalités (par exemple le SSO).

Nous avons bien précisé que nous ne disposions pas de l’énergie, des compétences en Go (le langage dans lequel est écrit Mattermost) ni (surtout) du temps nécessaire pour accepter les contributions externes, ou même les prestations qui nous étaient proposées.

Nous écrivons ce message afin de pouvoir discuter ici librement entre personnes/structures/institutions qui souhaitent ne pas se retrouver dans une impasse communautaire.

Nous souhaitons notamment faciliter la mise en contact entre :

  • personnes/structures/institutions qui seraient prêtes à financer un fork « solide » de Mattermost/Mostlymatter (notez, par ailleurs, que Framasoft est prête à mettre un peu d’argent, même si nous sommes une petite association à but non lucratif aux moyens limités)
  • personnes/structures/institutions qui seraient prêtes à maintenir et développer un tel fork (prestations rémunérées, donc)

Nous postons ce message en français et en anglais mais encourageons les réponses en anglais.

Hi,

at work I belong to a team that manages a Mattermost instance devoted to innovation management, in a French public structure. Our instance is roughly 60 times smaller than the Framasoft instance.

We just discussed this post, and while we are definitely eager to use and contribute to such a fork (with a focus on OIDC), we are at a loss on how to « do our share »:

  • paying would be difficult;

  • contributing code or tests is an option, but we can’t be the ones who produce the whole OIDC plugin.

I wonder how many are in the same situation as us, and what options are available to split the work in smaller chunks that can be addressed by more persons/organisations.

Ideally of course, some organization says « I’m going to lead the initiative on this plugin » and we can become contributors to that effort. Or one organization says « I’m going to pay » and another says « I can take the money and do it », and again we can become users or contributors. But what if there are dozens of us in the same situation and none able to take lead or pay? Does anyone have ideas on how decentralized work can occur here?

Hello,

Universities in Rhineland-Palatine also use (currently) Mattermost for a lot of projects and scientific discussion groups in a large scale, where the fact that it’s open source, reliable and sustainable is way more important than the question of price. So keeping a fully features open source version alive with no surprises of losing features over time, maintained by an active community, is a strong goal.

Framasoft has already done a great job by removing a constructed artificial limit and maintaining a patch for this, so I’d think that keeping Mostlymatter in the productive state that Mattermost has once been, even if no new features would be added, would be worth to be funded by all interest groups in education and research relying on souvereign platforms. If this is beyond the current development capabilities of Framasoft to add needed functionality of single signon and federated login in a fork continuously, any commercial Open Source development company “could” also create a fork of the last working version, and get paid for continuous maintenance on a contract base, but in my experience, it just works better if a truly open source-affine and vendor-independent community is the motor.

I’ll check back later with more information.

Regards

-Klaus

We at Obmondo would like to contribute to developing an SSO feature.. patch.. if Mattermostly is willing to take it?
We are 95% golang developers - and would do it the same way, we have done for our internal go APIs - which is using Keycloak for authentication (which is oauth2).

Would love to see this!

Hello Pyg,

Like others smalls non-profit associations, we can’t afford the licence for our instance, but we can contribute to the funding of a fork.

I don’t come to the forum often, so call me if you need to.

Antoine - Zoomacom