[standards-jig] IBB: Making it all "go"

Sebastiaan Deckers cbas at screaming3d.com
Thu Apr 10 19:08:38 UTC 2003

Rachel Blackman wrote:

>I think it's better that if this will be pushed forward and standardized
>officially, it should be pushed forward and standardized in the /proper/
>form.  Making an official IQ-based IBB will only encourage client authors
>in the future when it's deferred to add the IQ one as they run down the
>list of JEPs.  (Don't believe me?  How many clients still implement
>x:avatar?) ;)

The avatar JEP (#8) was written more than one and a half year ago and 
has received no follow up.
My conclusion, based on an implementation and public testing for about 
six months, is that the x:avatar protocol works well enough, despite one 
or two minor design flaws.  It's even being used on phones (fyi, the 
original complaints about the JEP were related to inefficient bandwidth 
usage, hah!).
Implementations of IBB using IQ will most likely lead to the same 
conclusion that it is *good enough* for users.  Now can't we all just 
stop with the pointless bickering over message vs. iq and write some 
implementations, so XMPP users see some progress?

>Peter Saint-Andre wrote:
>>3. Implement IBB with IQ in some clients and see how it works
>>Not everything in Jabber needs to go through the JEP process, and in
>>this instance it might be valuable to pursue #3 and gain some
>>experience with IBB before seeking ratification of the protocol.
>This seems like the ideal route to me.  For those (Tijl) who need it
>immediately, they can come up with an IBB using IQ, and implement it.
>Others who need it for some problem can implement it and experiment with
>it.  Maybe IQ turns out to work well, or maybe the every-packet-ack does
>end up being excessive.  Either way, there's no active-status JEP to
>mislead a new client author into blindly implementing that JEP when it
>might not be the final Jabber 1.0 route.
>It'd be nice to gain some experience and have that information come back to
>the standards list, since either way it would be useful to have return to
>the community.

Couldn't have said it better.  (But I tried anyway :-P)


More information about the Standards mailing list