XSF Discussion - 2020-02-23

MattJ: the performance in muc:prosody@ is why I predict Jabber will adopt the qline (refuse to federate with servers which do not refuse to federate with server X)

hi

https://bgzashtita.es/upload/2648c95f187eff879a53c9890338a1a8d74210e6/YEc60NzVWiVweLH1nIymi0xmPHyCEpboo2TyEWd5/Screenshot_20200219-235041.jpg

only testing.

If one wanted to transport Prometheus metrics over XMPP, would it make more sense to simply slap the text format of the metrics into an XML element as character data, or do you think that restructuring the data as XML makes more sense?

I tend towards the former, because it’ll be easier to interface that with existing Prometheus tooling ✏

I guess it depends on your use case. But since those two appear to not rule out each other, I'd probably start with the easiest approach

well, they rule each other out in the sense that nobody would want to support two different protocols for the same thing ;)

Is the data easy to transform? Key value pairs or similar?

Key-value pairs, tagged with a timestamp, another key-value map of parameters, a type, and a description.

How far from https://xmpp.org/extensions/xep-0039.html ?

Can anyone bring fippo back? I was going to ask him about status of http://hancke.name/jabber/jingle-rtcpmux.patch

seems like it stopped with this https://mail.jabber.org/pipermail/jingle/2013-May/001916.html

rion, consider sending him an e-mail, cc'ing standards@

I just dived into datachannels topic of jingle and was thinking how to properly maintain ICE connections, reduce amount of port allocation and reuse the ports from different jingle applications. And it seems fippo wrote all the related XEPs =)