Simon--

In the IMS Project we are using the "Tool Bash" to stimulate
implementations.  The name Tool Bash means "an effort to bash out
tools".  The emphasis is on a creating a working implementation in a
short time.  Sometimes the result itself is called a "Bash", i.e., "This
is a bash search tool."  Larger objectives and functionality may be
identified, but they are logically or progammatically "stubbed out".
The general model is Rapid Application Development.  A Tool Bash is a
group of people who get together expressly to create an implementation.
It is not part of IMS per se, as IMS is not in the business of creating
tools.  We do recognize the importance of tools, however, so we try to
assist those who are trying to implement the specification.

A Tool Bash involves a small number of people, 4-8, who get together
specifically to design and write code.  It is hosted by one
organization, but supported by the participants as best can be done.
Participants ("Bashers") come from several organizations, so the
objective of interoperability is intentionally addressed.  It is not a
formal group, but a small self-organized community.  The group is made
up of people; getting the right ones is important.  The balance within
the group is important: data base specialists, XML specialists, UI
specialists and so forth are needed.  The people all need to be the sort
who can work with others; this is not a place for prima donnas.  Bashers
need to be intelligent, knowledgeable, creative, and able to accept
others' ideas.  The work atmosphere is highly interactive, and ideas
need to flow freely and be respected.  Many good ideas come from the
collaborations.

The group sets up an electronic forum of some sort to discuss concepts
and architectures before meeting.  The meeting(s) run a full 5-day week
at one site, and can involve some late night sessions.  There are daily
show-and-tells to keep everyone running down the same path; it is
amazing how people working together closely can diverge if you don't do
this.  At the end of the meeting, tasks to be accomplished and sub-teams
to accomplish them are developed.  It's remarkably effective, hard work,
and actually a lot of fun.  Have an FTP site for backups and transfers;
when people are working long hours writing code, crashes will occur.
Count on it.  Beware of viruses.  Pizza is good.  As I lead the metadata
effort for the IMS, I've had the pleasure of working on a metadata Tool
Bash.

The hosting organization carries a disproportionate amount of
responsibility and cost.  Our Metadata Tool Bash was at Syracuse
University, hosted by GEM.  They were wonderful, and responsible for a
successful meeting.  One hopes the hosting organization gets a lot out
of the Bash.  They put a lot into it.

The IMS metadata specification is currently being harmonized with the
IEEE P1484 Learning Technology Systems Committee Learning Object
Metadata group's efforts; the ARIADNE project in Europe and IMS have
been working closely with many others to develop this standard.  The
IEEE P1484 LOM, and its implementations, support the Dublin Core through
explicit mappings to Dublin Core.

--Tom

Thomas D. Wason, Ph.D.
Metadata Lead
IMS Project: Defining the Internet Architecture for Education
(http://www.imsproject.org)
GEM: Gateway to Educational Materials (http://gem.syr.edu/)
[log in to unmask]
919.839.8187
-----Original Message-----
From: Simon Pockley <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: Thursday, June 17, 1999 7:33 PM
Subject: Re: Dublin ...sorry... Microsoft Core [Was Re: Microsoft Si


>If we can learn from the past,
>it is fairly sobering to look at the history of standards
>when it comes to character encoding.
>
>
>>From the 19th century, the pattern seems to be
>that almost as soon as well meaning groups established a standard
>along came a company that created its own solution, quickly implemented
it,
>and everyone followed.
>
>
>Effectively implementing DC so that anyone can use it
>requires a host of tools yet to be developed.
>If MS can create those tools and quickly integrate them into its
applications
>then they will probably control the standard.
>That such a loose knit group as the DC community
>has trouble creating application tools, because it acts through
concensus,
>may, sadly, turn out to be its undoing.
>Surely someone from MS is listening to this conversation
>and can suggest another way?
>
>
>Regards
>
>Simon
>
>
>
>
>
>
>
>
>--
>Dr Simon Pockley - Cinemedia
(http://www.cinemedia.net/FOD/FOD0050.html)
>3 Treasury Place Melbourne Australia [voice] 61 3 96511510
>[home] 61 3 94897905 [email] [log in to unmask]
>



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%