@frank
We tried to make changes to our fork so it is independent from NC Talk and hit roadblocks
The core of NC has hooks for Talk (i.e. spreed)
I don’t see anyone wanting to install multiple conferencing systems. So having Talk being installed with something else is not just realistic. And Talk has some cool features and integrations with files and calendar.
To be honest, a better alternative to all that we have a single NC Talk which supports both the internal signalling server and BBB / Jitsi. We already did the heavy-lifting for BBB. Can we just work with the developers of Talk and have multiple options for conferencing?
That should be the ideal option instead of this proliferation of conflicting apps.
You are definitely right. There must be a way to integrate Jitsi, BBB and may be others (open-source!) together with the actually foreseen, costly, signaling server connections.
Oh my god, a working videoconference integration within Nextcloud would be just awesome. I really think it should be within Nextcloud Talk, since it’s just very good for chat and this functionality would perfectly fit in the gap of not be able to videoconference and the “not reinventing the wheel” idea of nextcloud.
Great. Just to be clear. When I mentioned that is should be possible to install Talk and your app in parallel then I didn´t want to suggest that this makes sense for users. But it´s a good test that the code and namespaces of the apps don´t collide. Because this would be bad coding style and dangerous. If this is all fine then I suggest you should publish it at apps.nextcloud.com to get more feedback from users.
Now that the server code is released, I am not sure we need this effort anymore. I am thinking of shelving it.
But one thing is for sure, the community having started developing alternatives to Talk, has put pressure to open source the underlying server. So it is a success as we are all much better off now.
The challenge is that the code (especially PHP) makes use of the original Talk namespace. It would require a substantial amount of work to get it ready for the apps store (to avoid namespace conflicts).
It is a cool effort, but I only see a handful of people really interested.
i guess there are a lot interested… but not capable of doing the work.
and of course that’s a task that needs to be done.
but anyways… putting it away can’t be the solution… maybe if you’d give it out to the world “as it is” with some comments what would need to be done before releasing it officially?
We installed the finally open-sourced server and it works very well. It’s rather simple, a GO server that multiplexes Janus (which is a compiled into a binary). So it is very efficient. I believe we can host more attendees on it than Big Blue Button.
For the purposes of Nextcloud, I believe we still achieved our objective (by speeding up the release of the server) to get a full open-source conferencing solution. I think we can shelve this Talk-BBB effort. We continue to use BBB on our other learning product (https://www.classroomapp.com) as it is more geared towards education.