What we once had (at the height of the XMPP era of the Internet) (2023)
17 points by LolPython
17 points by LolPython
where you could actually chat from your Google Talk account with FB Messenger accounts, by using the XMPP bridge
This is your regular reminder that this false memory is mythological. It's a very common false memory, but it never happened. Facebook never federated, nor was their over even XMPP at the time so far as we know. They had a mediocre bridge for accessing FB chats from an XMPP client, but you absolutely could not chat to that from your Google Talk account.
There's a conflict of interest if singpolyma promotes it, but not me; he's one of the people behind jmp.chat, which is an XMPP bridge to the POTS network for $5/mo. You can also get a data SIM from them. I've been a satisfied customer for almost 5 years now; it's really nice to have a cell provider where the frontline support are also the backend engineers. :)
This is not off-topic to the thread either; I'm pretty sure they're growing and building a future where XMPP is a core technology for telecommunications. Maybe the "XMPP era" the article describes was a local maximum, but I bet not a global one.
at least for facebook text messaging i can state as a fact that i could connect with a regular xmpp client like pidgin gajim conversations
Indeed you could, as I said. But you could not chat to Facebook users from a Google or otherwise non-Facebook account.
connecting as a client is not the same as facebook and gtalk servers federating with each other
Snikket is pretty easy to setup and run. There are a couple of decent XMPP clients, but yeah, no one uses it anymore. You'd have an easier time getting someone on IRC, despite the gap in functionality.
I think there is a wide gap between "everyone uses it" and "nobody uses it", that people seem to ignore. My family use XMPP to communicate with each other, and they don't know or care about the protocol, they just know it as "Snikket" or "the family chat". It's a useful tool in its own right, and while it would be nice if everyone was on a single protocol, it has benefits even for small groups.
I have various XMPP affiliations, including being the founder of Snikket, so feel free to adjust for bias. However I built Snikket to solve real-life problems, and it solves them, and I regularly hear from happy users. It doesn't have to be all or nothing.
That's really great! On a whim, I went and poked around with podman and got a working setup on my tiny VPS. I'll kick the tires and see if I can find people to chat with.
Keep up the great work!
To me, what many seem to think was a high point for XMPP was a low point. If the only reason you used XMPP was to interface with chat systems run by megacorporations, is that really what we want? Sure, I won't stand in the way of creating bridges to proprietary networks and I would love to see every chat network in the world move to or be accessible from the Jabber network, but my usage of XMPP and the Jabber network doesn't need Google or Meta.
With XMPP and the federated Jabber network, I have control over who I chat with and how (free open source software running on my devices and theirs, identity not tied to a phone number, etc.). I want to bring more people in my direction more than I want to meet them where they're at (even though the latter can be an important step toward the former). I can send someone a Snikket invite, and (if they accept) they're now free from whatever proprietary system they were using when they want to talk with me. I don't need access to thousands or millions of strangers; I need access to a handful of friends and family members.
People talk about client quality, but I assure you the clients of today are much better than the clients of 2008. But you can also still use the clients from 2008 if you want! That's a strength. We do have a long way to go building clients that look the way people have come to expect, and there are lots of outdated guides out on the Web suggesting software and servers people probably shouldn't use anymore. None of that means XMPP is less than it was in 2008. It's only more!
My paycheck does come from a service built on XMPP and free open source software, so of course you can adjust for bias as MattJ suggested. We work closely together building new and exciting things on top of XMPP and bringing more and more people onto the Jabber network. There are some brilliant people working on making XMPP better and better, and it's the only chat protocol I use with regularity.
Interesting - hopefully, we will go back to the more open and protocol-, not client-, driven internet.
Out of curiosity though: is it something inherent in the Matrix protocol that requires that much resources or is rather an implementation issue?
I think trying to be "protocol driven" is what often kills these efforts. For a user the app is king. There are huge benefits to the app using an open protocol. There are benefits to their friends using different apps to be able to chat with them. But at the end of the day close contacts often want "the same" experience which of course requires using the same app, no matter what the protocol.
True; and it's also hard to build network effect around the protocol. Why apps should care to implement it, when nobody is there yet?
I used to host an XMPP server. It was nice. Some time after google defederated, it stopped feeling worthwhile to do that anymore. There were a few users directly on my server, but the really nice part was being able to chat with anyone who had gmail, with absolutely no fuss. (And with just a little fuss, you could use OTR with them for private chats.)
I also used to maintain a fork of the InstantBird XMPP client with a niche extension or two, before XUL got killed off. I still miss that one. My fork had a single digit audience, but it wasn't hard to maintain, and I liked it.