RFP: Remove reliance on public chat platforms #3527

Open
opened 2026-02-11 20:10:18 +01:00 by crueter · 9 comments
Owner

Issue

We rely very heavily on public chat platforms, to the point that it is pretty much impossible to get any testing or triaging done whatsoever without Discord or Revolt or whatever on God's green Earth it's called now. This is bad, and we should never rely on any platforms other than ones we control. Arguably, we should not rely on chat platforms in general, but public chat platforms is the main issue, especially given that Revolt and Discord are both pretty staunchly anti-free speech. The maintainership for Revolt is also extremely questionable, with a few faces I recognize who will bend over backwards to serve Microsoft or Discord or whomever.

I could go on an on about the millions of other problems like Discord's privacy disaster or the ID verification stuff (which is largely fearmongering but is a legitimate concern). But I'll keep that for later.

Problems to Solve

Some problems we would have to solve with a shift away from public chat platforms.

Testing Accessibility

Discord is "accessible" in the sense that everyone and their mother has an account, meaning becoming a tester is rather trivial. Of course, this means you have to use Discord to be a tester, which is a huge no. This problem should be solved regardless of if we choose to migrate.

Triaging

Triaging in general sucks, because I wake up every day with 5 trillion pings just saying "TRIAGE GOOD TO MERGE" without any context whatsoever. I don't really know what to do here other than encouraging people to check the Git for PRs that are ready to go.

Test Results

Just put them in the PR comments.

Community

I don't particularly care for community, but having a public-facing server to make showcases and announcements isn't a bad idea. We should eventually move to using our blog infrastructure for news, showcases, updates, etc. rather than... Discord.

Alternatives - Chat Platforms

If we want to stay on the live communication route.

revolt.chat

No.

Self-Hosted Revolt

Maybe. Revolt is a horribly made platform in general, with UX that was likely designed by a group of 12 year olds... not to mention the 15 different frontends that are all equally garbage.

Matrix

Federation with large public servers is an absolute no. Especially given that the official matrix.org instance has numerous connections to surveillance agencies like the CIA and Mossad.

XMPP

XMPP is great in general, because it doesn't use federation, users can use their own servers at-will, and it's private by default. Encryption is good too for small rooms. The barrier to entry is rather high, however, and making "servers" in the Discord sense isn't really supported (nor should it be, to be fair). The clients all suck, but whatever.

Alternatives - Testing

Separate App

Make an application that indexes currently available PR builds and have users download that, where they can submit feedback to a (self-hosted) server. Live communication may be feasible?

GitHub PR Discussions

Not really preferable, especially given the risk of DMCA.

Patreon

Unironically the best option.

### Issue We rely very heavily on public chat platforms, to the point that it is pretty much impossible to get any testing or triaging done whatsoever without Discord or Revolt or whatever on God's green Earth it's called now. **This is bad**, and we should never rely on any platforms other than ones **we** control. Arguably, we should not rely on *chat platforms* in general, but public chat platforms is the main issue, especially given that Revolt and Discord are both pretty staunchly anti-free speech. The maintainership for Revolt is also extremely questionable, with a few faces I recognize who will bend over backwards to serve Microsoft or Discord or whomever. I could go on an on about the millions of other problems like Discord's privacy disaster or the ID verification stuff (which is largely fearmongering but is a legitimate concern). But I'll keep that for later. # Problems to Solve Some problems we would have to solve with a shift away from public chat platforms. ## Testing Accessibility Discord is "accessible" in the sense that everyone and their mother has an account, meaning becoming a tester is rather trivial. Of course, this means you *have* to use Discord to be a tester, which is a huge no. This problem should be solved regardless of if we choose to migrate. ## Triaging Triaging in general sucks, because I wake up every day with 5 trillion pings just saying "TRIAGE GOOD TO MERGE" without any context whatsoever. I don't really know what to do here other than encouraging people to check the Git for PRs that are ready to go. ## Test Results Just put them in the PR comments. ## Community I don't particularly care for community, but having a public-facing server to make showcases and announcements isn't a bad idea. We should eventually move to using our blog infrastructure for news, showcases, updates, etc. rather than... Discord. # Alternatives - Chat Platforms If we want to stay on the live communication route. ## revolt.chat No. ## Self-Hosted Revolt Maybe. Revolt is a horribly made platform in general, with UX that was likely designed by a group of 12 year olds... not to mention the 15 different frontends that are all equally garbage. ## Matrix Federation with large public servers is an absolute no. Especially given that the official matrix.org instance has numerous connections to surveillance agencies like the CIA and Mossad. ## XMPP XMPP is great in general, because it doesn't use federation, users can use their own servers at-will, and it's private by default. Encryption is good too for small rooms. The barrier to entry is rather high, however, and making "servers" in the Discord sense isn't really supported (nor should it be, to be fair). The clients all suck, but whatever. # Alternatives - Testing ## Separate App Make an application that indexes currently available PR builds and have users download that, where they can submit feedback to a (self-hosted) server. Live communication may be feasible? ## GitHub PR Discussions Not really preferable, especially given the risk of DMCA. ## Patreon Unironically the best option.
crueter changed title from RFC: Remove reliance on public chat platforms to RFP: Remove reliance on public chat platforms 2026-02-11 20:13:53 +01:00
Author
Owner

This should go without saying, but put your proposals and thoughts HERE, not in Discord or Revolt.

This should go without saying, but put your proposals and thoughts HERE, not in Discord or Revolt.
Member

#80

https://git.eden-emu.dev/eden-emu/eden/issues/80
Member

Also unless there's a ping for Triage or anyone annoying, some PRs just get staled, see for example #3459@Discord, #3496, #2958, #2973

I will stop using @Triage altogether and bridge Testers reports as workaround for the time being

I dont have any proposal to make

Also unless there's a ping for Triage or anyone annoying, some PRs just get staled, see for example [#3459@Discord](https://discord.com/channels/1367654015269339267/1468363869423538318), #3496, #2958, #2973 I will stop using @Triage altogether and bridge Testers reports as workaround for the time being I dont have any proposal to make
Member
https://spacebar.chat/
Author
Owner

maybe

maybe
Author
Owner

Currently viable alternatives

In no particular order

these are mostly for if we absolutely have to self-host.

### Currently viable alternatives In no particular order - https://jami.net/ - https://townhall.chat/ - https://sharkord.com/ - https://shortwave.chat/ - https://spacebar.chat - https://github.com/fluxerapp/fluxer these are mostly for if we absolutely have to self-host.
Member

https://spacebar.chat/ seems viable

I would advise we stick to the main instance (yes that puts us at the mercy of the instance operators)
OTHERWISE
We should federate with the main instance at the very least, existing as a standalone would greatly hinder usability

https://spacebar.chat/ seems viable I would advise we stick to the main instance (yes that puts us at the mercy of the instance operators) OTHERWISE We should federate with the main instance at the very least, existing as a standalone would greatly hinder usability
Author
Owner

@DraVee wrote in #3527 (comment):

I will stop using @Triage altogether and bridge Testers reports as workaround for the time being

Pinging triage isn't the problem, it's the fact that we get pinged without context, and Triaging through... Discord is a bad idea. Things get lost super easily, and "staled" PRs get staled anyways.

@DraVee wrote in https://git.eden-emu.dev/eden-emu/eden/issues/3527#issuecomment-19786: > I will stop using @Triage altogether and bridge Testers reports as workaround for the time being Pinging triage isn't the problem, it's the fact that we get pinged *without context*, and Triaging through... Discord is a bad idea. Things get lost super easily, and "staled" PRs get staled anyways.
Author
Owner

We are going to move forward with spacebar now. Will need to look at what to do re: bots, but their API is similar so it should be possible to port it.

IFF we end up having to move off of federated/public instances, we can self-host this OR fluxer.

We are going to move forward with spacebar now. Will need to look at what to do re: bots, but their API is similar so it should be possible to port it. IFF we end up having to move off of federated/public instances, we can self-host this OR fluxer.
Sign in to join this conversation.
No milestone
No project
No assignees
3 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
eden-emu/eden#3527
No description provided.