Federation takes to the Web

[Jack Park]

I've been doing a lot of research related to http://wave.google.com/

Here is a snippet of what I think about Google Wave:

I see it as a step function (punctuated equilibrium in terms of
evolution) similar to that which the wiki platform introduced. Given
that it is open and open source, I believe it might become not only
useful, but quite possibly the way forward for some aspects of
knowledge federation, particularly the user interface aspect of a
federation (what ever that is).

I've spoken in terms of topic maps as an "interlingua" for federation;
regardless of which knowledge representation (KR) scheme someone uses
for their information resources, if there is a mapping into the
interlingua, then a federation can exists when that mapping is
performed. The other aspect of that federation is the creation and
evolution of those information resources. Wave provides an open,
extensible approach, all wrapped in a type of interlingua of its own:
an XML-based object called a "Blip". A Blip is the content object of
a Wavelet, which is a component of a Wave, which is a collaboratively
editable container object that lies at the heart of Google's Wave API
and protocol.

The protocol is a communications protocol based on the open standard
XMPP (Jabber). It provides a "real time" communications channel for a
given wave: if more than one browser is open to a particular wave,
say, one I opened and invited some friends to join and collaborate in
creating, then each individual sees the others typing in real time. A
demonstrated feature (see the 1.4 hour video) is that I can select
which language I want to compose in, and others will see that
composition in their language of choice. We can all create rich text
in international character sets (I'm waiting to see it handle Chinese
or Japanese translations in real time).

Where does federation come in? Wave servers are not restricted to
Google's servers. The server is to be open sourced and Google
encourages others to implement Wave servers of their own variety; so
long as the protocol is not violated, the servers exist as a
federation; I can invite you in from a server I authenticated in, say,
California, and you can view it from a server elsewhere. Any Wavelet
you create in my Wave will exist on your server, not mine. In the API
email list, this feature has been suggested to put a huge burden on
spammers; any spam they create resides only on their server.

Since the Wave API is extensible, you are able to create server-side
extensions, called Robots (Java or Python), and client-side
(javascript) features called Gadgets, which are based on the
OpenSocial.org gadget api. Thus, Waves can already contain opensocial
gadgets like stock tickers, weather forecasts, and more. Robots and
Gadgets can be written to support spreadsheets and more.

I happen to believe that it is a reasonable exercise to think about
ways in which Robots and Gadgets can be fabricated to support topic
mapping: federation inside a federation. Take a simple example, the
creation of a Concept Map (nodes with labeled arcs). Blips can have
child Blips. A Blip contains an XML string with elements, attributes,
and plain or styled text values. If a Blip element is, say, "node
type", then I could craft an IBIS dialogue tree by simply fabricating
an anchor Blip and growing a tree with child Blips, leaving labeled
arcs to the ontology underlying the particular IBIS dialect. If, in a
different scenario, I were to create a graph with either labeled arcs
of various kinds, or more to the topic mapping point, arcs that are
also addressable nodes, then, using the identity feature of a Blip
within a Wavelet, I might (I say "might" because I have yet to
actually try this) be able to generate Blips that reference other
Blips in a fashion similar to the way we now wire nodes in topic maps.
That remains to be seen.

Robots will be required to take apart the XML in a Blip; that, it
seems to me, means that a Wave might entail the equivalent of
"document types" which entail XML parsers and generators specific to
those types, thus the need to be able to find mappings among
documents, um, Waves.

I see this as a useful inquiry. I'd like to join with a few black-belt
developers and move this inquiry forward, and also with some HCI
black-belts to craft useful Gadgets for presentation and elicitation.

On a slightly different but still related concept, I suggest viewing
the 6 minute video about Primal Fusion at
http://corp.primalfusion.com/blog/?p=50

Cheers
Jack

[Dino Karabeg]

Nice!

I would just like to highlight a design principle they used: Instead of adhering to the metaphors we happen to live by because we grew up with them (the letter, the mailbox, the desktop, the file, the filing cabinet) and by doing that implementing in the new technology the patterns of use that evolved based on the old technology, they design a completely new object and patterns of use that are (1) efficient,  natural and fun from the point of view of the purpose they aim to fulfill and that (2) take advantage of the possibilities of the new technology. Then they make the new objects understandable and natural to people by finding a metaphor that reflects its own intrinsic power and principle of operation (the wave). 

The same design principle can be reapplied in a variety of  other contexts. 

An example is our Knowledge Federation development project.

We have been exchanging ideas and I believe we are converging to a rather similar strategy for (at least one part of) the future development of our project. 

This coming Thursday we will be dialoging about this strategy with the research group in the Trinity College Dublin where Hendrik is working. What follows is the last slide  of my introductory talk that points at the key insight, and its explanation (I will tell you the details of the strategy later, after they have been more carefully federated):

<attached picture>

Around the year 1950 the first computers came to the market and we developed computer science as the discipline and the profession whose task was to create algorithms, programming languages and everything else that was needed for programming this new device. 

Around the year 2000 the Web came and ... surely enough, we continued to think and act as computer scientists. We continued to  create computer applications.

But the Web allows us to do a lot more than that, because the Web is a lot more than a single computer – it allows us to 'program'  on the scale of the Web. The corresponding 'instructions' will involve both people acts and machine acts. An example of a resulting  'program'  may be a whole new social organization of the creation and communication of knowledge. (Of course, I am using 'program' and 'instruction' metaphorically, these concepts will acquire a whole new meaning in the new context.)

So think about a social organization of the creation and communication of knowledge that is (1) more effective, natural and fun and that (2) takes proper advantage of the available technology.

I will write more about this soon.

Warm greetings,

Dino 

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "KnowledgeFederation" group.
To post to this group, send email to knowledge...@googlegroups.com
To unsubscribe from this group, send email to knowledgefedera...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/knowledgefederation?hl=en
-~----------~----~----~----~------~----~------~--~---

[Jack Park]

What I am now beginning to see is a connection between Waves and
Professor Yuzuru Tanaka's Meme Media. His work may have contributed to
the wave concept, though I have yet to see references.

Jack

On Tue, Jun 2, 2009 at 12:35 PM, Dino Karabeg <dino.k...@gmail.com> wrote:
> Nice!
> I would just like to highlight a design principle they used: Instead of
> adhering to the metaphors we happen to live by because we grew up with them
> (the letter, the mailbox, the desktop, the file, the filing cabinet) and by
> doing that implementing in the new technology the patterns of use that
> evolved based on the old technology, they design a completely new object and
> patterns of use that are (1) efficient,  natural and fun from the point of
> view of the purpose they aim to fulfill and that (2) take advantage of the
> possibilities of the new technology. Then they make the new objects
> understandable and natural to people by finding a metaphor that reflects its
> own intrinsic power and principle of operation (the wave).
> The same design principle can be reapplied in a variety of  other contexts.
> An example is our Knowledge Federation development project.
> We have been exchanging ideas and I believe we are converging to a rather
> similar strategy for (at least one part of) the future development of our
> project.
> This coming Thursday we will be dialoging about this strategy with the
> research group in the Trinity College Dublin where Hendrik is working. What
> follows is the last slide  of my introductory talk that points at the key
> insight, and its explanation (I will tell you the details of the strategy
> later, after they have been more carefully federated):