4 ms·
> We don't know why only one of the big providers started interoperating with XMPP. But we do know why that one stopped again: A terrible wave of spammy contact
by MattJ100 6y ago
> We don't know why only one of the big providers started interoperating with XMPP. But we do know why that one stopped again: A terrible wave of spammy contact requests. They were able to fight intra-gtalk spam much more effectively than spam that originated via XMPP.
I think you're mistaken to believe this over the fact that Google was going through one of its turbulent internal transitions. The whole of Google Talk got replaced by Hangouts (which is now facing its own complex demise: https://arstechnica.com/gadgets/2019/01/the-great-google-hangouts-shutdown-begins-october-2019/ https://arstechnica.com/gadgets/2019/01/the-great-google-han... ) and the whole Google+ shift undoubtedly played a part.
Spam is not the problem. I have a Gmail account, and I receive tonnes of spam daily. I've had spam in Hangouts, I've had spam in Google Docs. Meanwhile independent and voluntary public XMPP server operators have quite a good handle on spam on the network today. I refuse to believe that Google's resources were unable to solve this, rather than that they were unwilling to solve this and took it as an excuse to wind down their participation in an open network.
I'm all for more research into anti-spam solutions, any open network has to deal with it. But so far none of the well established open networks (phone, email, physical mail) have yet solved the problem.
- Arnt 6y agoI may be mistaken, what I've heard is rumour and telling it violated someone's NDA, but I think the burden of proof is on your side. Consider Signal, since that's in the subject. Despite having my phone number on several web sites, I don't get spam on Signal, which I take to mean that the Signal maintainers are successful at fighting spam (and it doesn't matter whether they do so mostly as server operators or protocol designers). If you wanted to persuade them to interoperate/federate, then you have to provide arguments, ie. the burden of proof is on you. Saying "none of the well established networks have yet solved the problem" isn't enough, because you'll be trying to persuade someone who has solved the problem. You could try to argue that spam is a worthwhile price to pay, or that it could be fought as well using your protocol as via Signal's own servers, or other things. What you can not do is lean back. An interop mechanism is a server-to-server bulk message delivery mechanism, and so far I've not heard about any way to do bulk message delivery that doesn't atttract the interest of spammers. That spammers would try to spam Signal users via the interop mechanism is plausible by default, just as it's plausible that spammers would try to spam Gmail/gtalk users.