3 ms·

Somewhat tangential but relevant, what the hell happened to XMPP, period? Wasn't it supposed to be our one communication account we do groupchat, video calls, I

by zanny 11y ago

Somewhat tangential but relevant, what the hell happened to XMPP, period? Wasn't it supposed to be our one communication account we do groupchat, video calls, IMs with persistence, and file transfers over? Why did it just up and die with nobody working on it and nobody using it? http://xmpp.org/extensions/xep-0045.html http://xmpp.org/extensions/xep-0045.html exists. Its groupchat, it could probably be amended to support inline code, and its persistence model lets a server chose to broadcast the message history or not. And it is incredibly more user-friendly to make an XMPP account through a GUI client, have username@server, then join group@server for groupchat. No /join, no /register, no port management or #channel management. Hell, any reasonable XMPP client would default to your current server for groupchat so when you click "join chat room" you get the server part prefilled and you just enter the group name (or pick it from a list of groups - the protocol supports channel announcements). I get that XMPP is a horrible format (XML) and horribly bloated (committee) and horribly old (1999?) but if its that broken why can't we make another protocol for peer to peer realtime communications? Why isn't webrtc becoming that transport? Why do we keep getting these one shot web startups trying to replace open protocols every other weekend for a quick buck?

upofadown 11y ago
Well XMPP groupchat does in fact work (as you say). So the question would be; how does it compare to Slack? ... and IRC?
scrollaway 11y ago
Previous discussion on this: https://news.ycombinator.com/item?id=10280611 https://news.ycombinator.com/item?id=10280611 It's sad. But the only thing that can be done is work on an actual protocol that will solve these problems. It's a big burden to bear, and more often than not, building something that will bear such a burden just kills it outright (cf... xmpp.)
- ZenoArrow 11y ago
  > "But the only thing that can be done is work on an actual protocol that will solve these problems." What about the Matrix protocol (referenced in other comments made in this discussion)?
  - scrollaway 11y ago
    I am following it from afar. I don't see how it'll gain traction, ever, though. The marketing problem is one devs don't like to solve, but it has to be solved.
    - Arathorn 11y ago
      The plan for marketing it is: * Ensure there are glossy Matrix-native clients out there with a Slack-equivalent level of good UI and UX (e.g. http://vector.im/beta http://vector.im/beta) * Build bridges to federate together as many different protocols as possible. This includes IRC, Slack, XMPP, SIP, Lync, Verto, Respoke etc. We're trying to make this as easy as possible by providing some high level building blocks like http://github.com/matrix-org/matrix-appservice-bridge http://github.com/matrix-org/matrix-appservice-bridge. This hopefully will make it easy for others to bridge themselves into Matrix once we have critical mass. * Release a stable version of actual Matrix standard itself, keeping it as simple but capable as possible. (Right now it's not frozen, and it's still in beta) * Provide a really capable reference server implementation. (The current one, synapse, is a good starting point and works well for experimenting with Matrix, but we still have work to do on performance and scalability). * Run around the world encourage existing projects/networks/solutions/developers to get involved and defragment their comms - a bit like the early internet guys had to do with defragmenting email. It's obviously hugely ambitious, but we actually have some good traction already (despite still being in beta). Hopefully the trend will continue! [Disclaimer: i work on it]
      - scrollaway 11y ago
        Best of luck! I'll do my best to support you.
    - heavenlyhash 11y ago
      I got my mother to use Matrix. She's a smart cookie, granted. But she's not a developer. I couldn't reasonably get her to use IRC. I couldn't get her to use XMPP. I couldn't get her to use Facebook chat. (Moral reasons, there, mostly.) She could get started using Matrix in a single sitting. We're still using it to communicate. Anecdata? Absolutely. With a dash of emotional appeal, even -- the worst kind of anecdata! Still: true. So, while Matrix may not have a million dollar marketing budget in play... there's reasonable hope that the ease-of-use requirement for mass adoption is well on its way.

by zanny 11y ago

upofadown 11y ago

Well XMPP groupchat does in fact work (as you say). So the question would be; how does it compare to Slack? ... and IRC?

scrollaway 11y ago

Previous discussion on this: https://news.ycombinator.com/item?id=10280611 https://news.ycombinator.com/item?id=10280611 It's sad. But the only thing that can be done is work on an actual protocol that will solve these problems. It's a big burden to bear, and more often than not, building something that will bear such a burden just kills it outright (cf... xmpp.)

ZenoArrow 11y ago
> "But the only thing that can be done is work on an actual protocol that will solve these problems." What about the Matrix protocol (referenced in other comments made in this discussion)?
- scrollaway 11y ago
  I am following it from afar. I don't see how it'll gain traction, ever, though. The marketing problem is one devs don't like to solve, but it has to be solved.
  - Arathorn 11y ago
    The plan for marketing it is: * Ensure there are glossy Matrix-native clients out there with a Slack-equivalent level of good UI and UX (e.g. http://vector.im/beta http://vector.im/beta) * Build bridges to federate together as many different protocols as possible. This includes IRC, Slack, XMPP, SIP, Lync, Verto, Respoke etc. We're trying to make this as easy as possible by providing some high level building blocks like http://github.com/matrix-org/matrix-appservice-bridge http://github.com/matrix-org/matrix-appservice-bridge. This hopefully will make it easy for others to bridge themselves into Matrix once we have critical mass. * Release a stable version of actual Matrix standard itself, keeping it as simple but capable as possible. (Right now it's not frozen, and it's still in beta) * Provide a really capable reference server implementation. (The current one, synapse, is a good starting point and works well for experimenting with Matrix, but we still have work to do on performance and scalability). * Run around the world encourage existing projects/networks/solutions/developers to get involved and defragment their comms - a bit like the early internet guys had to do with defragmenting email. It's obviously hugely ambitious, but we actually have some good traction already (despite still being in beta). Hopefully the trend will continue! [Disclaimer: i work on it]
    - scrollaway 11y ago
      Best of luck! I'll do my best to support you.
  - heavenlyhash 11y ago
    I got my mother to use Matrix. She's a smart cookie, granted. But she's not a developer. I couldn't reasonably get her to use IRC. I couldn't get her to use XMPP. I couldn't get her to use Facebook chat. (Moral reasons, there, mostly.) She could get started using Matrix in a single sitting. We're still using it to communicate. Anecdata? Absolutely. With a dash of emotional appeal, even -- the worst kind of anecdata! Still: true. So, while Matrix may not have a million dollar marketing budget in play... there's reasonable hope that the ease-of-use requirement for mass adoption is well on its way.