User:Larma/Council Candidacy 2021

Jump to navigation Jump to search

Hi, This is my application to serve on XMPP Council in the 2021/22 term.

About me

Read about me on my user page. Not going to copy this over 😉

Council Candidacy 2021

I've never been a member of the council in the past, so this would be my first year. It's always good to shuffle people a bit and get new people aboard, right?

Accepting XEPs as Experimental

In my opinion, the hurdles for accepting a XEP as Experimental shouldn't be very high. It needs to pass formal requirements. It needs to define a goal. And it needs to define semantics and syntax sufficiently enough, that independent developers (that have prior experience with XMPP) can implement it and create compatible implementations (if applicable and only as much as needed to reach its goal). Notable it can have basically the same goal as existing other XEPs for as long as it is clear, what is different and why the author thinks it's a good idea. Also, an experimental XEP is not required to be compatible with other existing (or planned) XEPs - although compatibility or the intend to be compatible with widely implemented stable XEPs is of course desirable.

Mark XEPs as stable when they are

XEPs are stable when they are widely implemented, there is no common understanding that something needs to be changed ASAP and incompatible changes would not be possible without adversely affecting the ecosystem for years to come. This was, however, not really a criteria for marking XEPs as stable in the past, which is why many XEPs were not moved to stable. Of course there are formal requirements to be fulfilled for a XEP to move to stable, but we have to be honest with ourselves about when something actually is stable. In the end this can even lead to people developing a (maybe incompatible) replacement, thus bringing development forward faster. This means we'd end up in deprecating and obsoleting more prior stable standards, but I think that's a good thing.

If XEPs stay in Experimental (or Deferred) forever, we make it unnecessary hard for people to know if they are meant to be used as is, if there is active development around them, if one should propose changes or rather start something new if the current version does not suffice, etc.