[standards-jig] JEPs and Jabber Adoption

Tijl Houtbeckers thoutbeckers at splendo.com
Sun Jun 29 17:52:26 UTC 2003


Ryan Eatmon <reatmon at jabber.org> wrote on 29-6-2003 19:49:43:
>
>Linus maintains a kernel.  You are free to do whatever you want to 
>with your local copy of the kernel.  The JSF maintains a protocol.  
>You are free to ignore the JSF and DO WHATEVER you want to in your 
>Client. 
>
>No one ever said that in order to use XMPP to transport your data you 
>must have an accepted JEP.  Just like no one said you had to get 
>Linus' approval to write a new VM.  You just have to get everyone on 
>board when you want your thing to be accepted as a standard.

Currently this process in the JSF Jabber community is reversed. First 
get everyones approval, then make something, rather then make something 
(working with others, often out in the open and within the community), 
and have it approved. 

>I use Jabber to transport lots and lots of things.  Some of them have 
>been written up by JEPs, but I'm not putting forth my view of how they 
>should be done.  Instead I offered up my experience in comments to 
>improve the JEPs.  In the future, I might switch over to the standard, 
>but for now my homebrew protocol works jsut fine for me.
>
>And going through the process to be accepted as a standard can be 
>annoying.  That's my point.  You are free to do what you want when you 
>want.  Just don't expect that when you come forward with an idea to be 
>a standard that we aren't going to push back and tell you work with 
>others to merge ALL of the ideas into a usable form for everyone.

So rather then having competing ideas and implementations in an open 
fashing to see what works we end up trying to force different ideas 
together before even getting any work done, that have to compete with 
propriety solutions (jabber and non-jabber based). That doesn't always 
work that well (after all, why should you?) and not that much is gained 
from it either in the end. It does slow everything down and make 
everyone tired. 

I think a more open apporoach would spur jabber development, and we'll 
end up with a lot of good, useable, and broadly implemented protocols 
as a result. It would surthenly broaden the appeal of the jabber 
community. And I haven't heard one good reason why this would plunge us 
into a dark world where nothing interoperates. What seems to be a side-
effect of the current approach, that we're actually advising people to 
go and use their own propierity solution that will only work with 
*their* products cause we're incapable of producing a standard, worries 
me more in that respect. 

I think that'll be my last contribution to this discussion for a while.

-- 
Tijl Houtbeckers
Software Engineer @ Splendo
The Netherlands




More information about the Standards mailing list