My name is Julian K. Missig, but I am often known as x-virge. The rumors are true, I just graduated high school and will be attending Carnegie Mellon University to study psychology starting this fall. I joined Jabber around server release 0.7 to help out with the protocol and make Jer more correctly support XML Namespaces. Somehow I have managed to stick around ever since.
I spent quite a bit of time as a web site designer/developer, so design holds a special place in my heart. It was through my interest in XML that I found Jabber and helped correct some issues with compliance. I currently am employed by IBM as a Software Engineer.
My primary area of contribution to Jabber has been with the protocol. The past two years I have been better known for Gabber, which I helped Dave Smith create and then took over when he left. My name is on several JEPs, and I can be blamed for XHTML Basic finding its way into Jabber messages.
First and foremost, I would like to see the Browse v. Discovery issue resolved. We must pick one of the two and make it into the standard requirement for all Jabber nodes. I feel this is crucial for Jabber to move forward.
The publication/subscription problem should be resolved. I have seen great uses for publication/subscription, and things such as music information and avatars simply cannot be implemented properly without it. In order for Jabber to become accepted by any members of the general public, we do need to have some focus on features such as these.
After these two issues are resolved, the ideas formulated in the Standards Jabber Interest Group for namespace tables should be further investigated. Creating tables of which namespaces should be supported for a "Lightweight Jabber Client" versus a "Jabber Transport" will give implementors a starting guide and assist greater consistency across platforms.
While the current Jabber protocol is pretty good for what it does, the protocol must move forward. Simply getting up and writing a whole new protocol then expecting everyone to move to it is unrealistic, while retaining full backwards compliance will keep us back. I think we should take an approach somewhere in the middle—slowly incorporting the new ways of doing things while phasing out the older ways. This is the only way that I can see us keeping our current userbase while purging ourselves of early mistakes.
One of the first "new ways of doing things" I would like to push is proper support for XML Namespaces. My first steps with XML Namespaces were just to ensure that Jabber does not completely break the specification. However, this is far from good enough now that more and more applications are using XML. Parsers now exist which can properly handle namespace prefixes and all the other oddities of the full XML Namespace specification. Properly supporting XML Namespaces will help those who are coming from other XML communities use Jabber, and in turn help us gain support from those communities.