[humanmarkup-comment] Re: [pe-forum] planet-earth roadmap

[Rex Brooks <rexb@starbourne.com>]

A few suggestions, Viveka,

1-I would take the lead from the Web 3D Consortium, (of which I know 
you are completely aware) and put everything on an XML basis so that 
everyone will be able to use the the current and subsequent ISO 
Standards, including H-Anim, which I am working on, too.

2-http://www.starbourne.com/Inter3D.html--my contribution is a proof 
of concept world, without the mu/interactive functionality yet, but 
it inlines Mark Pesce's WebEarth, which seems like a pretty healthy 
starting point.

3-You might want to check with Michael Wagner, wagner@asu.edu and see 
if his associate, whose name I am currently not remembering beyond 
Casimir, (which I think is his given first name), who had a couple of 
Masters candidates working on a multimedia blob-type (if I remember 
correctly) database--a combined object-relational construct-- before 
settling on a database format, and then I think making sure it is 
interoperable with the open, public, if not open source, databases 
would also be a good idea. They both teach at Arizona State 
University.

4-I would use all of geovrml since it makes no sense to reinvent the 
wheel or to boil the ocean when you have a lot of available, on-going 
work to choose from. My own little world is not yet connected to 
geovrml since I have yet to put it in the next logical format--X3D 
plus geovrml plus h-anim plus multi-user and fully interactive.

No doubt you are already well aware of all of that except for perhaps 
my little neighborhood.

I suspect that if you ask, Mark Pesce would be happy to allow you to 
use the facilities for WebEarth.

As for myself, I will happily contribute the HumanMarkup code for a 
middleware driver for H-Anim when we get there. This framework will 
provide means for an on-demand, physics-based, behavioral engine 
which will be using HumanMarkup as the core for the facial 
expression/bodily gesture elements. There will probably be several 
developed as modules from which one will be able to choose, though I 
know of at least three under construction, all of which will be 
interoperable and have different application areas on which they 
focus.

If you are unaware of HumanMarkup, it is two pronged:

An entirely encumbrance-free, open, public, OASIS TC effort to write 
the standard specification:

http://www.oasis-open.org/committees/humanmarkup

and A 501 (c) (3) Non-Profit Corporation to support and promote 
education about the Human Markup Language Specifications and 
associated projects: Humanmarkup.org, Inc.:

http://www.humanmarkup.org

(Note: this website is in transition and so you are temporarily 
obliged to download the links but this will be corrected in due time 
with a full makeover.)

Everything I mention here is modular, xml-based-interoperable, and 
will provide the free, open, public substrate upon which for-profit 
plug-in modules can be based in order to provide for a means to 
stimulate applications and use, but which also provides for 
governmental, educational, and non-profit public service uses free of 
cost (and royalty-free) as well.

Ciao,
Rex

At 3:19 PM +1000 6/2/02, viveka wrote:
>Good hello. This is a message to the planet-earth mailing list; all 
>of you have subscribed at some time. This message may come as 
>something of a surprise, since the project has been dormant for some 
>time.
>
>It is now, however, awake again. How awake? Well, one message's 
>worth. To wake it up some more, reply.
>
>If you're no longer interested, you can unsubscribe from the list at:
>http://www.pairlist.net/mailman/listinfo/pe-forum
>or by contacting me.
>
>This message will be about how the planet-earth project will 
>develop; long-term and short-term goals, and some things we can do 
>immediately to get it moving.
>
>Long-term: the vision is still very much like the planet earth 
>application from snowcrash. We want a virtual representation of the 
>earth, updated in realtime from multiple sources, which can be 
>accessed in arbitrary ways, searched, annotated, filtered and 
>navigated.
>
>How do we build this? There are two main components; the database 
>and the interface.
>
>I'm an interface designer. Some of you will already be aware that I 
>can't code my way out of a wet paper bag. This has led me to 
>approach the problem from the wrong direction; specifying 
>interaction details rather than building core code. A look at the 
>documentation section of the website will show you this; if you have 
>Cosmo Player and you dig around, you can play a pretty VRML demo 
>that does *nothing at all*.
>
>I've spent the last couple of years learning about information 
>architecture. This will hopefully allow me to approach the problem 
>from the other direction; database design. Hopefully, this will be 
>more fruitful. (You'll notice that the middle part - actual *code* - 
>still escapes me :)
>
>I think I've gained a couple of key insights. One is that the 
>project isn't about representation; it's about providing an API. At 
>the core, it's about creating a standard data format for 
>geospatially-referenced information that people can easily use and 
>contribute their local data to, and providing a reference 
>implementation so they can start building the database.
>
>The other part is this: it must be radically decentralised. If it's 
>going to scale, it has to be end-to-end, like the Net itself. Anyone 
>should be able to set up a server, and to define their own data 
>elements. Anyone should be able to write a viewer, or even to 
>extract specific data from the database for use in another app, in 
>another context.
>
>Alright, before I get too much deeper, some things we can do now, to 
>get the project online. We have three projects:
>
>1. The database
>2. The HTML reference implementation
>3. The VRML reference implementation
>
>Steps for project 1:
>
>I. Define a basic element set. I suggest that we start with the Dublin Core:
>http://dublincore.org/documents/1999/07/02/dces/
>and the DCMI Point Encoding Scheme:
>http://dublincore.org/documents/dcmi-point/
>and the Box Encoding Scheme:
>http://dublincore.org/documents/dcmi-box/
>
>II. Build an implementation (probably in postgresql) of that database.
>
>Steps for project 2:
>I. Define a format for browsing the database.
>2. Write HTML code to read and write the database.
>
>Steps for project 3:
>1. Design an interface for navigating the database (done).
>2. Write VRML code to read and write the database.
>
>I suggest that we deploy the reference implementation at
>http://alpha.planet-earth.org
>Which does not yet exist.
>
>I've been looking at http://www-124.ibm.com/developerworks/oss/ as a 
>place to host it; does anyone have any thoughts about that? As 
>before, we'll license everything under the GPL.
>
>We have a CVS set up at sourceforge which we can start using once we 
>have a need for it.
>
>If I've missed anything, I won't be at all surprised. Please respond 
>with comments.
>
>OK, who's in, and for which bits?
>
>Regards,
>
>V.
>--
>Viveka Weiley, Karmanaut.
>{ http://www.karmanaut.com | http://www.planet-earth.org
>    http://www.MacWeb3D.org | http://sydney.siggraph.org.au }
>Hypermedia, virtual worlds, human interface, truth, beauty.
>
>_______________________________________________
>pe-forum mailing list
>pe-forum@planet-earth.org
>http://www.pairlist.net/mailman/listinfo/pe-forum


--