Open Source for privacy concerns
Open Source for privacy concerns
opensourceprivacy@cloud.schnuffle.de
Open Source for privacy concerns

  zuletzt bearbeitet: Thu, 19 Apr 2018 19:43:46 +0200  
IS XMPP dead?
Short story, from my point of view: YES!!

I've been runing XMPP servers for quite a while, be it prosody, ejabberd or openfire. It always ended the same way, after some time I dropped my server because I wasn't using it at all.
The problem with XMPP is the X in the standard, standing for "extensible". When talking about protocols, the ability of a protocol to be extended is seen as a strong point and I totally agree with that. But when you have a look at the history of XMPP, it has shown, that the formally strong point has lead to the exact opposite.
When the standard base is lacking many needed features which have to be added later on and the standardization process is captured by global corporate groups, the end result is a spec monster, that is more shaped by political and strategic decisions then by  technological need.

Nowadays a IM protcol needs following features at least:
* OS support for Windows, OSX, Linux
* Voice/Video support
* Mobile support (Android and IOS )
* E2E communication
* Group chats

There're dozens of XEPs that offer all needed functionality, so it shouldn't be a problem?
Apparently it still is, the only viable mobile XMPP client is Conversation, which is strong on the privacy side but totally lacks voice/video support. Try converting non IT people to use such a poor alternative and you will utterly fail ( at least that happened to me several times, so maybe that's an unfair generalization).

And even when we drop mobile, the overall experience isn't much better.  XMPP misses the flagship client application, the server documentations are mostly incomplete, there's no settled standard setup ( which XEPs supported, STUN/TURN integration, ...... ).

So for now I'm leaving XMPP, maybe I come back when things have settled in some years :).

My actual alternative playground, which by far isn't perfect is Matrix.org

Bild/Foto

An open network for secure, decentralized communication. An open network for secure, decentralized communication. Your support helps Matrix to build the decentralised communciations network of the future! If you use or build atop Matrix, or if you share our ideals of decentralisation, encryption, and open communication as a basic human right, yo...


They have some problems to overcome and I wouldn't bet my money on it yet, but it shows that with a protocol that incorporates the needed features from the beginning, a project can deliver a much better user experience.

When I reread my post, I realize that in the end my critic might be more about the change from technological driven standardization like it has been done at the W3C  to corporate driven  like OASIS or ISO.

Anybody remembers ECMA-376 aka Office Open XML File Formats?

I leave it here
have fun
Schnuffle

  zuletzt bearbeitet: Thu, 19 Apr 2018 19:45:11 +0200  
When I've been reflecting  on voice/video messaging, I always thought that to achieve optimal user experience the codecs need to be optimized for changing bandwidth.
It seemed so obvious that I was wondering that there isn't much info about such codecs available.
Now there's one promising research project that exactly does what seems so obvious, the Salsify project at Stanford University

Salsify — A New Architecture for Real-time Internet Video