[Foundation] position paper -- Mike Lin

Mike Lin mikelin at MIT.EDU
Thu Jul 25 00:19:25 CDT 2002


Position Paper -- Mike Lin for Jabber Council
email: mikelin at mit.edu
jid: mlin at xmpp.mit.edu
www: http://mikelin.mit.edu

For the past two years I've been straddling many areas where Jabber has
potential, as a Jabber Council member, a sophomore at MIT, and a
programmer for IBM. I believe this has put me in a pretty strategic
position to at all at once contribute to our collective understanding of
how Jabber fits in to the bazaar of the open source community, the ivory
tower of academia, and the profit-minded beuracracy of an international
megacorporation.

I'll now outline my contributions thusfar in each of these areas. As you
may have noticed, I tend to write few but long and verbose posts, such
as this one, on the topics that matter to me; I will link to them for
your reference, and I invite you to read through them for a better
understanding of my positions.

1. Bazaar of the Open Source Community

As a Jabber Council member, I have been primarily focused on two major
issues.

(a) Jabber (tm)

I have long pushed Jabber, Inc. to provide us with a clear and
unambiguous acceptable usage policy for the Jabber trademark. While some
laudable progress was made in this direction, I believe there is still
more that should be done - eventually leading us on the long-term road
to JSF ownership of the trademark. Stemming perhaps from my experience
with intellectual property issues in my other two areas of contribution,
it is my opinion that many underestimate the importance of fair and
non-profit administration of this central trademark.

http://mailman.jabber.org/pipermail/members/2002-March/000804.html
http://mailman.jabber.org/pipermail/members/2002-March/000943.html

(b) Next Generation Jabber Transport

>From my perspective, there are two major things the next generation
Jabber transport should achieve: (1) massive scalability and
manageability, and (2) significant differentiation from SIP and related
technologies. I have worked (and argued) with many of you in the
open-source community towards these goals. With the help of others, I
have developed a lot of ideas -- and have run a lot of tests -- in this
area in search of the best way to move structured information in large
quantities between abstract nodes. By the way, I don't think BEEP is the
answer.

http://mailman.jabber.org/pipermail/standards-jig/2002-June/001006.html
http://www.jabber.org/jeps/jep-0017.html
http://mailman.jabber.org/pipermail/standards-jig/2002-May/000848.html

2. Ivory Tower of Academia

I have thusfar told only a few of you about work I've been doing for the
past several months in developing Jabber frameworks for the high-level
programming languages Scheme and Objective Caml. These languages are
widely popular in academic circles, and in my opinion they provide
vastly superior linguistic faculties for XML processing. I think
easy-to-use Jabber constructs for these languages would be a first step
in getting Jabber in use for academic purposes.

I will be rolling out some of the work I've been doing in the next few
weeks.

More broadly, I think Jabber, especially with a pub/sub engine, can
provide a powerful messaging substrate for "widely distributed"
computing applications research. While there is much work to be done to
make this realistic, and it faces intimidating competition from powerful
suites like Ensemble
(http://www.cs.cornell.edu/Info/Projects/Ensemble/), I believe that if
Jabber enjoys sufficient success in other areas, and if we work out
solutions to security and authentication, its ubiquity will make it a
compelling choice for distributed computing research.

3. Profit-Minded Beauracracy of an International Megacorporation

I've been an on-and-off part-time and full-time programmer at IBM
Internet Technology since just before I became involved with Jabber two
years ago. I contribute to the design and development of IBM Sash
Weblications for Windows (http://sash.alphaworks.ibm.com/), an
e-business client application development technology. I have developed a
complete suite of Jabber runtime capabilites for Sash, which allows Sash
Weblications to message and receive alerts via Jabber; although I'm not
sure if anyone outside of IBM uses it, this stack includes centralized
profile management and other desktop integration features not, to my
knowledege, available in any other Jabber stack.

I also developed a full-featured Jabber instant messenger in Sash,
uncreatively entitled SashJab. Although I'm not sure if anyone outside
of IBM uses it, this client has included for more than a year things
like emoticons, wizards, and an extensive help system that, to my
knowledge, have only recently begun to appear in a few other clients.

http://sash.alphaworks.ibm.com/sashjab

Recently, our group developed a Sash instant messenger that tightly
integrates with IBM's internal Lotus Sametime service and LDAP
directory. This client is currently deployed to some 15,000 IBM
employees. We are currently working to augment the software with support
for external IM services - through Jabber.

--

This e-mail has not been brief because I have a lot to say about Jabber.
Even still, I think this provides only a glimpse of where I see Jabber
going - a universal messaging substrate for the transport of structured
information in virtually any context, unfettered by corporate interests
and intellectual property constraints. In line with the directions I
outlined above, my agenda as a Jabber Council member, if I am
re-elected, will include:

1) Continuing to work towards a resolution of the Trademark quesiton.
2) Development of a next-generation transport protocol
3) Pub/sub and message queueing
4) Authentication and trust, a subset of security considerations

In any case, expect to hear more from me in the coming weeks, as my work
on all three fronts continues.

Love,
Mike




More information about the Members mailing list