forgejo/discussions
49
43

The future of communication in the Forgejo space #375

Open
opened 2025年07月29日 11:17:07 +02:00 by Beowulf · 31 comments
Member
Copy link

There are concerns regarding matrix approaching.

Concerns:

  • Open-Core:
    We state on our website:

    Guaranteed 100% Free Software

    Forgejo will always be Free and Open Source Software. Furthermore we exclusively use Free Software for our own project development.

    But at the same time Synapse is developed by Element which also sells a Pro version: https://element.io/server-suite/pro
    There were also already PRs which were blocked because they implement "enterprise features": https://toot.kif.rocks/@davebloggt/114901852199969595

  • Premium accounts:
    Matrix introduced on their "flagship" server freemium and preemium accounts. Freemium accounts are limited an can e.g. no longer create public rooms. This is a concern esp. for the moderation accounts which are on matrix.org

  • ...

There are concerns regarding matrix approaching. Concerns: - Open-Core: We state on our website: > Guaranteed 100% Free Software > > Forgejo will always be Free and Open Source Software. Furthermore we exclusively use Free Software for our own project development. But at the same time Synapse is developed by Element which also sells a Pro version: https://element.io/server-suite/pro There were also already PRs which were blocked because they implement "enterprise features": https://toot.kif.rocks/@davebloggt/114901852199969595 - Premium accounts: Matrix introduced on their "flagship" server freemium and preemium accounts. Freemium accounts are limited an can e.g. no longer create public rooms. This is a concern esp. for the moderation accounts which are on matrix.org - ...

Ever since I can remember using Matrix, they were always something to dislike about Matrix because of something that happened in the last few months; I cannot recall a large time period where there wasn't some pressing issue (or drama) surrounding Matrix. I do not want to claim that contentiously developing and improving a (public) instant messaging protocol is easy, but I do feel that the problems that Matrix have faced and are still facing have not been technically motivated1 .

I do not go out of my way to see how Matrix operates and I simply do not care that much as long as I can happily chat away with people, I would even consider myself a 'normal user', I happily use matrix.org as my homeserver (despite that fueling the centralized nature of Matrix) and I only ever used Element (X) as my Matrix clients. That said, how I am looking at Matrix the foundation has changed drastically the last year.

I've seen quite a lot of posts (excluding the cannot decrypt message memes) in my Mastodon timeline that from my perspective have valuable critique about the Matrix foundation, ecosystem and protocol. The times that I check in a few days later and there's a official response by Matrix (or Element in the cases that's relevant) I am simply left unsatisfied with the reasoning.

In the last year most of these critiques have been about changes that have directly impacted many Matrix users2 . I do not have the trust that this is going to stop, as such I think it's good that the Forgejo project starts thinking and looking to leave Matrix as its primary way of communication.

  1. I would be one of the first to say this might be a human scaling problem, but knowing that there's a for-profit is footing the bill of Matrix and that they have a team of people I am inclined to think they do not have the scaling issue. Context: https://lobste.rs/s/vqsran/matrix_is_cooked#c_jekh0n ↩︎

  2. The somewhat recommended mobile app Element X is nowhere near feature parity with the original Element app and the UI is not usable if you are in many rooms. The room upgrades that has to be coordinated in advance and causes disruptions. Self-hosting Matrix homeserver has become time-involved. ↩︎

Ever since I can remember using Matrix, they were always something to dislike about Matrix because of something that happened in the last few months; I cannot recall a large time period where there wasn't some pressing issue (or drama) surrounding Matrix. I do not want to claim that contentiously developing and improving a (public) instant messaging protocol is easy, but I do feel that the problems that Matrix have faced and are still facing have not been technically motivated[^1]. [^1]: I would be one of the first to say this might be [a human scaling problem](https://blog.codeberg.org/the-hardest-scaling-issue.html), but knowing that there's a for-profit is footing the bill of Matrix and that they have a team of people I am inclined to think they do not have the scaling issue. Context: https://lobste.rs/s/vqsran/matrix_is_cooked#c_jekh0n I do not go out of my way to see how Matrix operates and I simply do not care that much as long as I can happily chat away with people, I would even consider myself a 'normal user', I happily use matrix.org as my homeserver (despite that fueling the centralized nature of Matrix) and I only ever used Element (X) as my Matrix clients. That said, how I am looking at Matrix the foundation has changed drastically the last year. I've seen quite a lot of posts (excluding the cannot decrypt message memes) in my Mastodon timeline that from my perspective have valuable critique about the Matrix foundation, ecosystem and protocol. The times that I check in a few days later and there's a official response by Matrix (or Element in the cases that's relevant) I am simply left unsatisfied with the reasoning. In the last year most of these critiques have been about changes that have directly impacted many Matrix users[^2]. I do not have the trust that this is going to stop, as such I think it's good that the Forgejo project starts thinking and looking to leave Matrix as its primary way of communication. [^2]: The somewhat recommended mobile app Element X is nowhere near feature parity with the original Element app and the UI is not usable if you are in many rooms. The room upgrades that has to be coordinated in advance and causes disruptions. Self-hosting Matrix homeserver has become time-involved.

This is a very sad turn of events 😢

This is a very sad turn of events 😢
Owner
Copy link

I only partially agree to the concerns as stated by Beowulf in the first message. Matrix is a protocol, and similar to email where there are proprietary or "open core" email servers and clients, the same can be true for Matrix. It is probably the only truly decentralized chat protocol. As far as I understand, not even XMPP would survive the server hosting a room going down. Matrix (in theory) would.

However, my primary concern with Matrix is the tight integration (also with regards to $$$) between for-profit and the Matrix Foundation. If the email protocol was under the hands of those that would promote open core and proprietary email servers, I wouldn't trust that protocol any longer. With Matrix, it seems that there is a critical conflict of interest between the Matrix foundation depending on the money from Element, and then developing the whole protocol and ecosystem into a direction that makes Element happy. IMHO the Matrix foundation does not act in the interest of the general public at all.

I a fully agree to Gusted's message. I have suffered from Matrix a lot and it's not a pleasant experience.

My own problems with the Matrix ecosystem

General:

  • The protocol is too complicated. I recently wanted to implement an imap to matrix tool and gave up trying to be compliant to the protocol. Instead, I chose to be only compliant to the clients I use. The protocol requires you to use a lot of fallbacks for clients that don't fully support the protocol and thus make it more difficult to implement the protocol. It's a circle. It might have been better to have a simple protocol that can be reasonably supported by every client, without the need to add complex fallbacks, because it is nearly impossible to support all of the protocol.
  • Lack of proper clients, probably due to the complex protocol. I tried nheko, Element (+X), Cinny, FluffyChat, Thunderbird and some others. None of them work well. Some are not maintained and simply broken. Element is super slow.
  • The Matrix foundation moves the protocol in a direction where only Synapse and Element can effectively follow (also read this part of a nice blog post)

Performance:

  • Element is super slow. Element X is "up to 6000x faster than any other Matrix client." (source). Still, it's the slowest chat app I have ever seen, so either they completely screwed up, or the protocol simply doesn't allow to develop a client and you can only choose between "very very slow" and "absolutely unusable". The fastest Matrix client on mobile was nheko (had it on a Pinephone with Mobian). Not on Android, though.
  • My Element X recently broke entirely. It indicates that sync has finished, but no new messages appear. I have to force-stop the app on Android and restart to make new messages appear. Sometimes, I can send messages in rooms, but they disappear (visually) from the client where I sent them. They only re-appear after similarly force-stopping and restarting the client. It 100% prevents me from participating in Forgejo conversations from mobile.
  • Element on desktop takes several minutes to load on my machine. Until it is usable, I might have forgotten what I actually wanted to do.
  • Element on desktop sometimes peaks at CPU usage seemingly at random. It prevents me from doing useful work, and I have to stop the client. Then it takes ages to start again. It kinda also prevents me from participating online.
  • nheko is somewhat usable for me. Only when resuming from standby, it takes a long time to reconnect, but otherwise works fine. It seems somewhat unmaintained, though, and I don't know if it will work with the upcoming room version 12.

UI/UX:

  • Element has tons of funding, and still they don't get simply UX patterns right. One example: Do you agree that "Reject" is the recommended (green) action when accepting a room invite failed? How many people are likely to click the wrong button (especially, because "Accept" was green and in this place before it failed) A red "Retry" and a green "Reject" button makes it likely for users to click on the reject button.
  • Element clients still nudge everyone to register on matrix.org instead of showing homeservers with equal priority
  • Many people I know never receive notifications from Element on mobile, even if the built-in notification test shows no problem. Especially if they are not regular users, they will likely miss out on important aspects, making Matrix unsuitable for occasional contributors that might forget to check back regularly.
  • All the confusion between Element and Element X, where the latter is beta and not near-where feature-parity with Element, but on the other hand it is promoted since Element is kinda considered legacy. It makes it really hard for "normal" users to even get started. Element (normal) cannot register an account on matrix.org from Android, for instance. You need Element X for that. But Element X doesn't have spaces and threads and is broken (see above).
I only partially agree to the concerns as stated by Beowulf in the first message. Matrix is a protocol, and similar to email where there are proprietary or "open core" email servers and clients, the same can be true for Matrix. It is probably the only truly decentralized chat protocol. As far as I understand, not even XMPP would survive the server hosting a room going down. Matrix (in theory) would. However, my primary concern with Matrix is the tight integration (also with regards to $\$$) between for-profit and the Matrix Foundation. If the email protocol was under the hands of those that would promote open core and proprietary email servers, I wouldn't trust that protocol any longer. With Matrix, it seems that there is a critical conflict of interest between the Matrix foundation depending on the money from Element, and then developing the whole protocol and ecosystem into a direction that makes Element happy. IMHO the Matrix foundation **does not act in the interest of the general public** at all. I a fully agree to Gusted's message. I have suffered from Matrix a lot and it's not a pleasant experience. <details><summary>My own problems with the Matrix ecosystem</summary> General: * The protocol is too complicated. I recently wanted to [implement an imap to matrix](https://codeberg.org/eventbike/imap2matrix) tool and gave up trying to be compliant to the protocol. Instead, I chose to be only compliant to the clients I use. The protocol requires you to use a lot of fallbacks for clients that don't fully support the protocol and thus make it more difficult to implement the protocol. It's a circle. It might have been better to have a simple protocol that can be reasonably supported by every client, without the need to add complex fallbacks, because it is nearly impossible to support all of the protocol. * Lack of proper clients, probably due to the complex protocol. I tried nheko, Element (+X), Cinny, FluffyChat, Thunderbird and some others. None of them work well. Some are not maintained and simply broken. Element is super slow. * The Matrix foundation moves the protocol in a direction where only Synapse and Element can effectively follow (also read [this part of a nice blog post](https://nexy.blog/2024/12/16/my-experience-on-matrix/#sliding-sync-deprecation-lack-of-respect-for-the-ecosystem-from-element)) Performance: * Element is super slow. Element X is "up to 6000x faster than any other Matrix client." [(source)](https://element.io/blog/element-x-experience-the-future-of-element/). Still, it's the slowest chat app I have ever seen, so either they completely screwed up, or the protocol simply doesn't allow to develop a client and you can only choose between "very very slow" and "absolutely unusable". The fastest Matrix client on mobile was nheko (had it on a Pinephone with Mobian). Not on Android, though. * My Element X recently broke entirely. It indicates that sync has finished, but no new messages appear. I have to force-stop the app on Android and restart to make new messages appear. Sometimes, I can send messages in rooms, but they disappear (visually) from the client where I sent them. They only re-appear after similarly force-stopping and restarting the client. It 100% prevents me from participating in Forgejo conversations from mobile. * Element on desktop takes several minutes to load on my machine. Until it is usable, I might have forgotten what I actually wanted to do. * Element on desktop sometimes peaks at CPU usage seemingly at random. It prevents me from doing useful work, and I have to stop the client. Then it takes ages to start again. It kinda also prevents me from participating online. * nheko is somewhat usable for me. Only when resuming from standby, it takes a long time to reconnect, but otherwise works fine. It seems somewhat unmaintained, though, and I don't know if it will work with the upcoming room version 12. UI/UX: * Element has tons of funding, and still they don't get simply UX patterns right. One example: Do you agree that "Reject" is the recommended (green) action when accepting a room invite failed? How many people are likely to click the wrong button (especially, because "Accept" was green and in this place before it failed) ![A red "Retry" and a green "Reject" button makes it likely for users to click on the reject button.](/attachments/836883d2-c061-4473-9b8f-55be0fbc96d3) * Element clients still nudge everyone to register on matrix.org instead of showing homeservers with equal priority * Many people I know never receive notifications from Element on mobile, even if the built-in notification test shows no problem. Especially if they are not regular users, they will likely miss out on important aspects, making Matrix unsuitable for occasional contributors that might forget to check back regularly. * All the confusion between Element and Element X, where the latter is beta and not near-where feature-parity with Element, but on the other hand it is promoted since Element is kinda considered legacy. It makes it really hard for "normal" users to even get started. Element (normal) cannot register an account on matrix.org from Android, for instance. You need Element X for that. But Element X doesn't have spaces and threads and is broken (see above). </details>
Author
Member
Copy link

@fnetX wrote in #375 (comment):

Matrix is a protocol, and similar to email where there are proprietary or "open core" email servers and clients, the same can be true for Matrix.

Yes, this is true, the focus is on can be true. The issue is that synapse is the reference implementation and the most deployed solution. And is developed by Element/The Matrix foundation.
But you worded the issue/concerns better 👍

@fnetX wrote in #375 (comment):

However, my primary concern with Matrix is the tight integration (also with regards to $) between for-profit and the Matrix Foundation.

Yes, there is no real distance between both.

@fnetX wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-6035074: > Matrix is a protocol, and similar to email where there are proprietary or "open core" email servers and clients, the same can be true for Matrix. Yes, this is true, the focus is on **can be true**. The issue is that synapse is the reference implementation and the most deployed solution. And is developed by Element/The Matrix foundation. But you worded the issue/concerns better 👍 @fnetX wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-6035074: > However, my primary concern with Matrix is the tight integration (also with regards to $$$) between for-profit and the Matrix Foundation. Yes, there is no real distance between both.
Owner
Copy link

Does it make sense to consider launching an open letter to the Matrix Foundation? I feel like there is a lot of concerns from the FLOSS community about the future of Matrix. And I believe the FLOSS community is probably the most important consumer of Matrix. Not because they pay for the service, but because they promote the protocol among FLOSS communities, other projects and likely even help with lobbying in the public sector. Or they sit in IT departments of companies and promote Matrix. This value is hard to measure, but I believe that Matrix would be much more difficult to promote in enterprise context if the FLOSS community was not involved.

Does it make sense to consider launching an open letter to the Matrix Foundation? I feel like there is a lot of concerns from the FLOSS community about the future of Matrix. And I believe the FLOSS community is probably the most important consumer of Matrix. Not because they pay for the service, but because they promote the protocol among FLOSS communities, other projects and likely even help with lobbying in the public sector. Or they sit in IT departments of companies and promote Matrix. This value is hard to measure, but I believe that Matrix would be much more difficult to promote in enterprise context if the FLOSS community was not involved.
Author
Member
Copy link

@fnetX wrote in #375 (comment):

Does it make sense to consider launching an open letter to the Matrix Foundation?

I think it could make sense if it is a joined effort (even though I don't think it will lead to a success). Maybe we can ask some other projects, e.g. postmarketOS to join?

@fnetX wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-6036937: > Does it make sense to consider launching an open letter to the Matrix Foundation? I think it could make sense if it is a joined effort (even though I don't think it will lead to a success). Maybe we can ask some other projects, e.g. postmarketOS to join?

Received that direct message, probably because I'm admin on some rooms in the Forgejo space.

We have updated our Homeserver Terms and Privacy Policy. We strongly encourage you to read these documents in full, but for clarity these are some of the main changes:

  • Updated the minimum age requirements for use of the Matrix.org Homeserver to be 18 years old;
  • Introduced new measures to comply with our obligations under the Online Safety Act and the Digital Services Act;
  • Introduced new payment terms to support paid plans on the Matrix.org Homeserver;
  • Describe the new data processors to support paid plans on the Matrix.org Homeserver.

Each of the documents has a detailed version history which we encourage you to review. The updated Homeserver Terms and Privacy Policy take effect on 7 August, 2025. These terms apply to you by continuing to use the homeserver after that date. If you have any questions please drop us an email to legal@matrix.org

Received that direct message, probably because I'm admin on some rooms in the Forgejo space. --- > We have updated our [Homeserver Terms](https://matrix.org/legal/terms-and-conditions/) and [Privacy Policy](https://matrix.org/legal/privacy-notice/). We strongly encourage you to read these documents in full, but for clarity these are some of the main changes: > > - Updated the minimum age requirements for use of the [Matrix.org](http://matrix.org/) Homeserver to be 18 years old; > - Introduced new measures to comply with our obligations under the Online Safety Act and the Digital Services Act; > - Introduced new payment terms to support paid plans on the [Matrix.org](http://matrix.org/) Homeserver; > - Describe the new data processors to support paid plans on the [Matrix.org](http://matrix.org/) Homeserver. > > Each of the documents has a detailed version history which we encourage you to review. The updated Homeserver Terms and Privacy Policy take effect on 7 August, 2025. These terms apply to you by continuing to use the homeserver after that date. If you have any questions please drop us an email to [legal@matrix.org](mailto:legal@matrix.org)
floss4good changed title from (削除) The future of communication in the forgejo space (削除ここまで) to The future of communication in the Forgejo space 2025年07月30日 13:00:58 +02:00
Author
Member
Copy link

Sure, they need money. But the question is: at what cost and why?

I think they themselves are mostly responsible for this situation. It's a federated network, but at the same time, if you open an official client (e.g. Element) or elsewhere the matrix.org home server is prominently advertised. This certainly leads to more people registering on morg. Mainly because a lot of people are new to this and just want to register to try it out and use it. There's a lot more I could write, but I think it would be offtopic because it would be more about matrix's decisions than forgejo's use of matrix.

Sure, they need money. But the question is: at what cost and why? I think they themselves are mostly responsible for this situation. It's a federated network, but at the same time, if you open an official client (e.g. Element) or elsewhere the matrix.org home server is prominently advertised. This certainly leads to more people registering on morg. Mainly because a lot of people are new to this and just want to register to try it out and use it. There's a lot more I could write, but I think it would be offtopic because it would be more about matrix's decisions than forgejo's use of matrix.

I've been using Matrix for 10 years now, and while many of the criticisms are certainly valid, I disagree with the characterization of it being open core or not fully committed to being an open solution.

We run our own Synapse server (more than one in fact), and have integrated a lot of systems with it. In my experience it has been quite a bit easier to manage and integrate with than XMPP or IRC. As a communication layer for the Internet along the lines of SMTP, I think they have done a lot to invite participation from the community and make it as open as possible. You can certainly argue about their effectiveness but I don't buy any argument that this is an open-core cynical ploy to profit off the commons.

I think Matrix is the best platform for communications out there, and we should all be holding them accountable in helping grow it in a way that benefits everyone -- not tearing them down for struggling to be sustainable.

If they won't accept a PR due to it conflicting with their paid version, there's nothing that stops any of us from forking it and managing that fork -- just like ForgeJo did with Gitea. Should we all switch to Gitlab?

Codeberg has terms of service, just like Matrix.org -- if you don't like them, nothing stops you from running your own ForgeJo server, just like nothing stops you from running your own Synapse (and the Matrix team encourages people to do exactly that, and has from the very start).

In other words, the very same criticisms people level at Matrix can be leveled at any open source project. And based on them setting up an independent foundation, I'd say they are better than Gitea ever was at being a responsible steward of open source -- there's plenty to criticize and nitpick, but cut them a little slack -- what else would you switch to, Discord?

I've been using Matrix for 10 years now, and while many of the criticisms are certainly valid, I disagree with the characterization of it being open core or not fully committed to being an open solution. We run our own Synapse server (more than one in fact), and have integrated a lot of systems with it. In my experience it has been quite a bit easier to manage and integrate with than XMPP or IRC. As a communication layer for the Internet along the lines of SMTP, I think they have done a lot to invite participation from the community and make it as open as possible. You can certainly argue about their effectiveness but I don't buy any argument that this is an open-core cynical ploy to profit off the commons. I think Matrix is the best platform for communications out there, and we should all be holding them accountable in helping grow it in a way that benefits everyone -- not tearing them down for struggling to be sustainable. If they won't accept a PR due to it conflicting with their paid version, there's nothing that stops any of us from forking it and managing that fork -- just like ForgeJo did with Gitea. Should we all switch to Gitlab? Codeberg has [terms of service](https://codeberg.org/Codeberg/org/src/branch/main/TermsOfUse.md), just like Matrix.org -- if you don't like them, nothing stops you from running your own ForgeJo server, just like nothing stops you from running your own Synapse (and the Matrix team encourages people to do exactly that, and has from the very start). In other words, the very same criticisms people level at Matrix can be leveled at any open source project. And based on them setting up an independent foundation, I'd say they are better than Gitea ever was at being a responsible steward of open source -- there's plenty to criticize and nitpick, but cut them a little slack -- what else would you switch to, Discord?

The open letter has not been written or even planned to be written, the criticism so far written is out of a personal capacity. If a open letter is happening it would be a new discussion and then feel free to post your thoughts about it - as for now this is mostly about Forgejo's future w.r.t. using which communication platform.

The open letter has not been written or even planned to be written, the criticism so far written is out of a personal capacity. If a open letter is happening it would be a new discussion and then feel free to post your thoughts about it - as for now this is mostly about Forgejo's future w.r.t. using which communication platform.
Member
Copy link

I am really not sure why the Monthly Updates are driving people that don't have much of a relation to the project to use threads like this one and #366 to interact with highly contentious issues involving how the project works in a manner that is not related to problem-solving and may even be potentially aggravating (in #366, although perhaps unintentionally) – please excuse me for being blunt but we're trying to organize ourselves here, it'd be best not to turn our place of work into a debate club or a Reddit comments section.

I am really not sure why the Monthly Updates are driving people that don't have much of a relation to the project to use threads like this one and https://codeberg.org/forgejo/discussions/issues/366 to interact with highly contentious issues involving how the project works in a manner that is not related to problem-solving and may even be potentially aggravating (in #366, although perhaps unintentionally) – please excuse me for being blunt but we're trying to organize ourselves here, it'd be best not to turn our place of work into a debate club or a Reddit comments section.
Member
Copy link

I'd describe the vibe here as "imagine that the folks working on Forgejo in some way or another are in a coffee shop and having a conversation". Sure, the conversation is taking place in public and you can join in. But, should you wish to get yourself involved and add something, please be try to be careful and courteous.

There might be a lot of implicit context (such as "there is no concrete proposal for an open letter yet, that's just how we talk here") that you might not be aware of – such misunderstandings can get disruptive quickly.

I'd describe the vibe here as "imagine that the folks working on Forgejo in some way or another are in a coffee shop and having a conversation". Sure, the conversation is taking place in public and you can join in. But, should you wish to get yourself involved and add something, please be try to be careful and courteous. There might be a lot of implicit context (such as "there is no concrete proposal for an open letter yet, that's just how we talk here") that you might not be aware of – such misunderstandings can get disruptive quickly.

I'm not trying to sow any arguments here. While I haven't been very involved in ForgeJo so far, I've been advocating for open source software for over 20 years. I'm a core maintainer of LedgerSMB and I contribute a lot to Drupal. I'm less involved but highly supportive of Matrix, NextCloud, ForgeJo, Home Assistant among other open source projects. I often blog about these, when I have something to say.

As a business owner, I use more open source software than probably any other American business. I tend to prefer copyleft licenses to BSD/Apache/etc, to protect projects. And while I think Matrix's struggles are very much related to taking VC funding to grow quickly, on the whole I trust their team to do the best they can. I hate to see projects leaving Matrix. So my recommendation, if there's a desire to stay on an open platform, is to run your own Matrix server. Doesn't even have to be Synapse - Conduit seems like it could be good enough.

I know that's a chunk of new responsibility to take on -- but with the long list of contributors, maybe this is something someone could take on? Perhaps even me...

I'm not trying to sow any arguments here. While I haven't been very involved in ForgeJo so far, I've been advocating for open source software for over 20 years. I'm a core maintainer of LedgerSMB and I contribute a lot to Drupal. I'm less involved but highly supportive of Matrix, NextCloud, ForgeJo, Home Assistant among other open source projects. I often blog about these, when I have something to say. As a business owner, I use more open source software than probably any other American business. I tend to prefer copyleft licenses to BSD/Apache/etc, to protect projects. And while I think Matrix's struggles are very much related to taking VC funding to grow quickly, on the whole I trust their team to do the best they can. I hate to see projects leaving Matrix. So my recommendation, if there's a desire to stay on an open platform, is to run your own Matrix server. Doesn't even have to be Synapse - Conduit seems like it could be good enough. I know that's a chunk of new responsibility to take on -- but with the long list of contributors, maybe this is something someone could take on? Perhaps even me...

Good morning,

Apologies in advance for the drift.

I've run across this thread in my search for the "Forgejo community." This has been difficult because Forgejo is a community based project so most search hits relate to that.

From what I read in this thread, Matrix is one way for developers (users?) to gather and chat. From a previous post "... coffee shop and having a conversation." I remain ignorant of Matrix, but that does not sound like what I'm looking for. What I hope to find is something searchable and which has answers to questions I have about using Forgejo. I'm active on Reddit, but it looks like there is very little activity there.

Where should I be looking?

Here are the questions I have at the moment:

  • Is there an existing script or program that can automagically mirror all repos on one site to another site? I see APIs that seem like they would be sufficient to accomplish this, but I didn't find anything that sounded like it would do that at delightful-forgejo
  • I have a repo with a lot of notes in Markdown format. I have a fairly kludgey way to publish these and I'd like to implement something better. I'd like to discuss what other users have done to accomplish similar. (I've explored the Wiki and it seems not suitable due to limits on directory depth and inability to link between pages.)

These are more user than developer oriented questions. And I'm unlikely to be the first to ponder them. I really prefer to leverage the efforts of those who have gone before me than reinvent the wheel.

Thanks for reading and if this is too far off topic, either the issue or repo owner or I can remove it and there will be no hurt feelings.

best,

Good morning, Apologies in advance for the drift. I've run across this thread in my search for the "Forgejo community." This has been difficult because Forgejo is a community based project so most search hits relate to that. From what I read in this thread, Matrix is one way for developers (users?) to gather and chat. From a previous post "... coffee shop and having a conversation." I remain ignorant of Matrix, but that does not sound like what I'm looking for. What I hope to find is something searchable and which has answers to questions I have about using Forgejo. I'm active on Reddit, but it looks like there is very little activity there. Where should I be looking? Here are the questions I have at the moment: * Is there an existing script or program that can automagically mirror all repos on one site to another site? I see APIs that seem like they would be sufficient to accomplish this, but I didn't find anything that sounded like it would do that at [delightful-forgejo ](https://codeberg.org/forgejo-contrib/delightful-forgejo) * I have a repo with a lot of notes in Markdown format. I have a fairly kludgey way to publish these and I'd like to implement something better. I'd like to discuss what other users have done to accomplish similar. (I've explored the Wiki and it seems not suitable due to limits on directory depth and inability to link between pages.) These are more user than developer oriented questions. And I'm unlikely to be the first to ponder them. I really prefer to leverage the efforts of those who have gone before me than reinvent the wheel. Thanks for reading and if this is too far off topic, either the issue or repo owner or I can remove it and there will be no hurt feelings. best,
Author
Member
Copy link

@HankB wrote in #375 (comment):

Where should I be looking?

The official room for questions is the matrix chat room #forgejo-chat:matrix.org.
If yoy don't want to use matrix, you could also try to ask your question in e.g. the fediverse.

@HankB wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-6218992: > Where should I be looking? The official room for questions is the matrix chat room #forgejo-chat:matrix.org. If yoy don't want to use matrix, you could also try to ask your question in e.g. the fediverse.

In case folks haven't seen this I wanted to share the video of Matthew from two weeks ago in which he addresses lots of concerns: YouTube: Matrix Live S11E07 - Reflections on Matrix criticism over the last week

In case folks haven't seen this I wanted to share the video of Matthew from two weeks ago in which he addresses lots of concerns: [YouTube: Matrix Live S11E07 - Reflections on Matrix criticism over the last week](https://youtu.be/OyuqM7RbX5E)
Author
Member
Copy link

Thanks for the link @moritzdietz . I had already seen the video, but hadn't watched it yet.

What I find very difficult at the moment is that the security issue, which has now been resolved, has been known since 2022 and that the fix was put on hold due to a lack of financial resources.
I understand that they are not in a "good" financial position, but in my opinion, security should be the top priority and other things should have been put on hold instead.

With this in mind, how can we be sure that there are no critical security vulnerabilities in Matrix that are known but not being fixed?

Work on fixing known cases of state resets in the current State Resolution 2.0 algorithm began in 2022 when David Robertson on Element’s backend team investigated known occurrences of the problem that were happening on the public network. He made good progress on identifying the root causes of these occurrences, resulting in the State Resolution v2.1 algorithm. Unfortunately, lack of funding meant work had to be paused.

https://matrix.org/blog/2025/08/project-hydra-improving-state-res/

Thanks for the link @moritzdietz . I had already seen the video, but hadn't watched it yet. *** What I find very difficult at the moment is that the security issue, which has now been resolved, has been known since 2022 and that the fix was put on hold due to a lack of financial resources. I understand that they are not in a "good" financial position, but in my opinion, security should be the **top priority** and other things should have been put on hold instead. With this in mind, how can we be sure that there are no critical security vulnerabilities in Matrix that are known but not being fixed? > Work on fixing known cases of state resets in the current State Resolution 2.0 algorithm began in 2022 when David Robertson on Element’s backend team investigated known occurrences of the problem that were happening on the public network. He made good progress on identifying the root causes of these occurrences, resulting in the State Resolution v2.1 algorithm. Unfortunately, lack of funding meant work had to be paused. https://matrix.org/blog/2025/08/project-hydra-improving-state-res/

Thanks for the link @moritzdietz . I had already seen the video, but hadn't watched it yet.

👍

I think it would make sense to reach out to the foundation and to get answers for something so delicate as this topic.
I'm not sure if the wording of the blog post meant "funding was an issue to tackle state res issues but now that we had, we uncovered this security vulnerability" or "back in 2022 we saw this security flaw in the State Res protocol but have had no money to fix it until now". Because this is a huge difference and leaving this up in the air free to interpret seems dangerous.

> Thanks for the link @moritzdietz . I had already seen the video, but hadn't watched it yet. 👍 I think it would make sense to reach out to the foundation and to get answers for something so delicate as this topic. I'm not sure if the wording of the blog post meant "funding was an issue to tackle state res issues but now that we had, we uncovered this security vulnerability" or "back in 2022 we saw this security flaw in the State Res protocol but have had no money to fix it until now". Because this is a huge difference and leaving this up in the air free to interpret seems dangerous.

I've explored Matrix using the Web API https://app.element.io/#/room/#forgejo-chat:matrix.org and using NeoChat. Matrix looks like a great way to chat with other users but I find it unsatisfying to explore possible solutions to my questions.

  • I'd be very surprised if any of my questions have not been asked and answered before. I don't see any way to search for that.
  • If there are several discussions in progress, all communications are intermingled making it difficult to follow one thread.

Perhaps I'm missing something. The kind of platform I would find useful is something like https://discourse.practicalzfs.com/. I find Reddit useful for some topics but I understand why some users have left and the Forgejo sub is not very active.

Are discussions on Codeberg encouraged or discouraged? I don't know how searchable they are but they are well threaded.

Barring that, I fear that I won't really be able to take full advantage of what Forgejo provides and see that as unfortunate considering how much effort goes into providing it.

best,

I've explored Matrix using the Web API <https://app.element.io/#/room/#forgejo-chat:matrix.org> and using NeoChat. Matrix looks like a great way to chat with other users but I find it unsatisfying to explore possible solutions to my questions. * I'd be very surprised if any of my questions have not been asked and answered before. I don't see any way to search for that. * If there are several discussions in progress, all communications are intermingled making it difficult to follow one thread. Perhaps I'm missing something. The kind of platform I would find useful is something like <https://discourse.practicalzfs.com/>. I find Reddit useful for some topics but I understand why some users have left and the Forgejo sub is not very active. Are discussions on Codeberg encouraged or discouraged? I don't know how searchable they are but they are well threaded. Barring that, I fear that I won't really be able to take full advantage of what Forgejo provides and see that as unfortunate considering how much effort goes into providing it. best,

Seconding a forum solution. not because of anything specific to Matrix (though my first impressions of it are pretty terrible) but that for someone who wants support on hosting an instance, an asynchronous and searchable communications method like a forum is much better than a synchronous one that requires a log in to even read.

Seconding a forum solution. not because of anything specific to Matrix (though my first impressions of it are pretty terrible) but that for someone who wants support on hosting an instance, an asynchronous and searchable communications method like a forum is much better than a synchronous one that requires a log in to even read.

I don't think a forum can replace the communication in the Forgejo space, I would feel this would mainly be for the main chatroom but all the special topic chatrooms would not benefit from a forum. I think at most it will be a addition to a instant messaging platform where longer discussions can be held.

I don't think a forum can replace the communication in the Forgejo space, I would feel this would mainly be for the main chatroom but all the special topic chatrooms would not benefit from a forum. I think at most it will be a addition to a instant messaging platform where longer discussions can be held.

Zulip seems to serve some projects in a middle ground that is both usable for real-time chat, and more complicated long-post forum discussions. Example adopter, Rust: https://rust-lang.zulipchat.com/

But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. 😒

Zulip seems to serve some projects in a middle ground that is both usable for real-time chat, and more complicated long-post forum discussions. Example adopter, Rust: https://rust-lang.zulipchat.com/ But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. 😒

@Gusted wrote in #375 (comment):

all the special topic chatrooms would not benefit from a forum

I think they could benefit a lot. For example, if I wanted to look up past discussions on the crawler incidents, I would find them split across days, and I would have to scroll past various topics sprinkled in between.
In a forum, these discussions would be in their dedicated topics and therefore would not interrupt each other, even across days or weeks of ongoing messages. Even more, parts of that discussion also happened in the discussions repo, so another place to search.

I also wouldn't be surprised if, currently, some discussions happen in different channels than they should (e.g. in Development instead of UI) because each channel needs to be joined explicitly (480 members in Development vs only 100 in UI), whereas different categories on a single forum would still be covered by a single account and a single search bar (and moving discussions would not create two separate streams).

@Gusted wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7151122: > all the special topic chatrooms would not benefit from a forum I think they could benefit a lot. For example, if I wanted to look up past discussions on the crawler incidents, I would find them split across days, and I would have to scroll past various topics sprinkled in between. In a forum, these discussions would be in their dedicated topics and therefore would not interrupt each other, even across days or weeks of ongoing messages. Even more, parts of that discussion also happened in the `discussions` repo, so another place to search. I also wouldn't be surprised if, currently, some discussions happen in different channels than they should (e.g. in Development instead of UI) because each channel needs to be joined explicitly (480 members in Development vs only 100 in UI), whereas different categories on a single forum would still be covered by a single account and a single search bar (and moving discussions would not create two separate streams).
Author
Member
Copy link

@mfenniak wrote in #375 (comment):

But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. 😒

I need to say, I find Zulip interesting (esp the threading/wiki like style) but never used it.
As far as I know Zulip isn't open core. And yes the notification system is not great. I also don't know if Zulip supports e.g. UnifiedPush, or only FCM...

The biggest disadvantage is that everyone needs to create an account just to participate in the forgejo space. But this would also e.g. the case with a forum. It is a big advantage of Matrix that user don't need to create seperate accounts...

@mfenniak wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7151173: > But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. :unamused: I need to say, I find Zulip interesting (esp the threading/wiki like style) but never used it. As far as I know Zulip isn't open core. And yes the notification system is not great. I also don't know if Zulip supports e.g. UnifiedPush, or only FCM... The biggest disadvantage is that everyone needs to create an account just to participate in the forgejo space. But this would also e.g. the case with a forum. It is a big advantage of Matrix that user don't need to create seperate accounts...

@mfenniak wrote in #375 (comment):

specifically the mobile notifications limitations in self-hosted open source systems

Apparently, open source projects can apply for a Community Plan ("which includes unlimited push notifications and support for many Zulip features").

@Beowulf wrote in #375 (comment):

everyone needs to create an account just to participate in the forgejo space

When self-hosted, OAuth/OpenID would alleviate that. I've had Discourse instances connected to an application feel like they're one system due to accounts being shared between them.

@mfenniak wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7151173: > specifically the mobile notifications limitations in self-hosted open source systems Apparently, open source projects can apply for a [Community Plan ("which includes unlimited push notifications and support for many Zulip features")](https://zulip.com/help/self-hosted-billing#free-community-plan). @Beowulf wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7195312: > everyone needs to create an account just to participate in the forgejo space When self-hosted, OAuth/OpenID would alleviate that. I've had Discourse instances connected to an application feel like they're one system due to accounts being shared between them.
Author
Member
Copy link

@sclu1034 wrote in #375 (comment):

Apparently, open source projects can apply for a Community Plan ("which includes unlimited push notifications and support for many Zulip features").

Yep hosting would theoretically also be free. The question is, if this would be in line with the values...

@sclu1034 wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7196083: > Apparently, open source projects can apply for a [Community Plan ("which includes unlimited push notifications and support for many Zulip features")](https://zulip.com/help/self-hosted-billing#free-community-plan). Yep hosting would theoretically also be free. The question is, if this would be in line with the values...

https://news.ycombinator.com/item?id=23106954 suggests they might. And the HN user name fits the top most contributor at github.com/zulip/zulip.

https://news.ycombinator.com/item?id=23106954 suggests they might. And the HN user name fits the top most contributor at [github.com/zulip/zulip](https://github.com/zulip/zulip).

@Beowulf wrote in #375 (comment):

As far as I know Zulip isn't open core. And yes the notification system is not great. I also don't know if Zulip supports e.g. UnifiedPush, or only FCM...

Hm... well, let me clarify my thought here as I dropped the term "open core" without being specific, and the link I posted doesn't reflect what I was looking at (changing between "Cloud" and "Self-hosted" plan tabs on https://zulip.com/features/ doesn't reflect in the URL 😛).

If you run a self-hosted Zulip server, Zulip uses a cloud-based push notification service that requires a subscription or an eligible Community plan to support more-than-10 users. https://zulip.readthedocs.io/en/stable/production/mobile-push-notifications.html As far as I can see the push notification service is not Open Source, as I can't find it on their GitHub page.

I think there are good technical reasons for centralizing this, as APNS and GCM require significant effort to set-up. And if Forgejo looked into using Zulip, I think the Community plan, either self-hosted or cloud-hosted, would probably be a viable option. But if Zulip goes out of business or revokes their community plans in the future, even if the service was self-hosted on the theoretical last Open Source release, it seems like there would be a partial functionality loss.

@Beowulf wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7195312: > As far as I know Zulip isn't open core. And yes the notification system is not great. I also don't know if Zulip supports e.g. UnifiedPush, or only FCM... Hm... well, let me clarify my thought here as I dropped the term "open core" without being specific, and the link I posted doesn't reflect what I was looking at (changing between "Cloud" and "Self-hosted" plan tabs on https://zulip.com/features/ doesn't reflect in the URL 😛). If you run a self-hosted Zulip server, Zulip uses a cloud-based push notification service that requires a subscription or an eligible Community plan to support more-than-10 users. https://zulip.readthedocs.io/en/stable/production/mobile-push-notifications.html As far as I can see the push notification service is not Open Source, as I can't find it on their GitHub page. I think there are good technical reasons for centralizing this, as APNS and GCM require significant effort to set-up. And if Forgejo looked into using Zulip, I think the Community plan, either self-hosted or cloud-hosted, would probably be a viable option. But if Zulip goes out of business or revokes their community plans in the future, even if the service was self-hosted on the theoretical last Open Source release, it seems like there would be a partial functionality loss.

That's going to be the story for mobile push-notifications for any service, though. APNS and GCM want a signed recipient, so either you go through a single central server, or you build and publish your own version of the app.

Long-polling the instance directly works, theoretically, but has negative effects on battery life, as you need to keep the app running.

Matrix isn't any better in this. My self-hosted element-web also needs to send notifications to a central gateway for the app to receive them. So if Vector goes out of business, I'm not getting push notifications for our current communications.

That's going to be the story for mobile push-notifications for any service, though. APNS and GCM want a signed recipient, so either you go through a single central server, or you build and publish your own version of the app. Long-polling the instance directly works, theoretically, but has negative effects on battery life, as you need to keep the app running. Matrix isn't any better in this. My self-hosted element-web also needs to send notifications to a central gateway for the app to receive them. So if Vector goes out of business, I'm not getting push notifications for our current communications.
Author
Member
Copy link

@mfenniak wrote in #375 (comment):

https://zulip.readthedocs.io/en/stable/production/mobile-push-notifications.html As far as I can see the push notification service is not Open Source, as I can't find it on their GitHub page.

In the link you shared:

The code for the push notification forwarding service is 100% open source and available as part of the Zulip server project on GitHub (specifically, here).

But missing UnifiedPush is (at least for me) a major blocker...

@mfenniak wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7218253: > https://zulip.readthedocs.io/en/stable/production/mobile-push-notifications.html As far as I can see the push notification service is not Open Source, as I can't find it on their GitHub page. In the link you shared: > The code for the push notification forwarding service is 100% open source and available as part of the [Zulip server project on GitHub](https://github.com/zulip/zulip) (specifically, [here](https://github.com/zulip/zulip/tree/main/zilencer)). But missing UnifiedPush is (at least for me) a major blocker...

@Beowulf wrote in #375 (comment):

@mfenniak wrote in #375 (comment):

But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. 😒

I need to say, I find Zulip interesting (esp the threading/wiki like style) but never used it. As far as I know Zulip isn't open core.

Throwing this in here: https://cinny.in in case you haven't seen this client yet.

@Beowulf wrote in #375 (comment):

The biggest disadvantage is that everyone needs to create an account just to participate in the forgejo space. But this would also e.g. the case with a forum. It is a big advantage of Matrix that user don't need to create seperate accounts...

Especially now that with Matrix's upcoming (or even already using) OIDC auth thing where you can use any other social login you have reduces the burden of having yet another account even further.

Especially for new users trying to join the Forgejo Matrix Space and rooms simplifying the onboarding experience. As with all things – things take time and nothing is perfect but that's currently out there being used.

Discourse also offers this login capability – I've been using my GH account to log into several forums for quite a while now.

Reading the previous comments I get the feeling that other alternatives have similar downsides and don't offer that many upsides in comparison. The initial posts concerns seem not really addressed in the past comments other than 7 thumps up on sending a letter to the foundation voicing concerns. I bet if you shoot Matthew an E-Mail or DM him, he'll get in touch with you.

I say this very explicitly as I know from the past that there was a community that had issues and they never once reached out to the Matrix Foundation to voice those concerns directly and find a way to either resolve them/some of them or make plans on how to resolve them in the future.

For me personally I can say that I really appreciate that I can hang out with the Ansible and Mozilla people and at the same time being able to see what shenanigans the Fedora, Foreman, Forgejo or Woodpecker community is up to from within the same environment and on my own homeserver. There is much value in this.

No doubt: information gets lost in the scroll back buffer. This is where a forum shines – I feel like the Codeberg discussions space strikes a good balance between a forum and a somewhat ephemeral chat.

I write this with no spite or some mind game – saying this just in case it comes across this way :D

@Beowulf wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7195312: > @mfenniak wrote in #375 (comment): > > > But its 'Open Core' model might be a blocker for this community... https://zulip.com/features/ specifically the mobile notifications limitations in self-hosted open source systems. :unamused: > > I need to say, I find Zulip interesting (esp the threading/wiki like style) but never used it. As far as I know Zulip isn't open core. Throwing this in here: https://cinny.in in case you haven't seen this client yet. @Beowulf wrote in https://codeberg.org/forgejo/discussions/issues/375#issuecomment-7195312: > The biggest disadvantage is that everyone needs to create an account just to participate in the forgejo space. But this would also e.g. the case with a forum. It is a big advantage of Matrix that user don't need to create seperate accounts... Especially now that with Matrix's upcoming (or even already using) OIDC auth thing where you can use any other social login you have reduces the burden of having yet another account even further. Especially for new users trying to join the Forgejo Matrix Space and rooms simplifying the onboarding experience. As with all things – things take time and nothing is perfect but that's currently out there being used. Discourse also offers this login capability – I've been using my GH account to log into several forums for quite a while now. Reading the previous comments I get the feeling that other alternatives have similar downsides and don't offer that many upsides in comparison. The initial posts concerns seem not really addressed in the past comments other than 7 thumps up on sending a letter to the foundation voicing concerns. I bet if you shoot Matthew an E-Mail or DM him, he'll get in touch with you. I say this very explicitly as I know from the past that there was a community that had issues and they never once reached out to the Matrix Foundation to voice those concerns directly and find a way to either resolve them/some of them or make plans on how to resolve them in the future. For me personally I can say that I really appreciate that I can hang out with the Ansible and Mozilla people and at the same time being able to see what shenanigans the Fedora, Foreman, Forgejo or Woodpecker community is up to from within the same environment and on my own homeserver. There is much value in this. No doubt: information gets lost in the scroll back buffer. This is where a forum shines – I feel like the Codeberg discussions space strikes a good balance between a forum and a somewhat ephemeral chat. I write this with no spite or some mind game – saying this just in case it comes across this way :D
Sign in to join this conversation.
No Branch/Tag specified
No results found.
No results found.
Labels
Clear labels
User research - Accessibility
Requires input about accessibility features, likely involves user testing.
User research - Blocked
Do not pick as-is! We are happy if you can help, but please coordinate with ongoing redesign in this area.
User research - Community
Community features, such as discovering other people's work or otherwise feeling welcome on a Forgejo instance.
User research - Config (instance)
Instance-wide configuration, authentication and other admin-only needs.
User research - Errors
How to deal with errors in the application and write helpful error messages.
User research - Filters
How filter and search is being worked with.
User research - Future backlog
The issue might be inspiring for future design work.
User research - Git workflow
AGit, fork-based and new Git workflow, PR creation etc
User research - Labels
Active research about Labels
User research - Moderation
Moderation Featuers for Admins are undergoing active User Research
User research - Needs input
Use this label to let the User Research team know their input is requested.
User research - Notifications/Dashboard
Research on how users should know what to do next.
User research - Rendering
Text rendering, markup languages etc
User research - Repo creation
Active research about the New Repo dialog.
User research - Repo units
The repo sections, disabling them and the "Add more" button.
User research - Security
User research - Settings (in-app)
How to structure in-app settings in the future?
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
11 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Reference
forgejo/discussions#375
Reference in a new issue
forgejo/discussions
No description provided.
Delete branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?