[Embedlets-dev] Embedlets Group's Strategy

[Ted K. <tk...@ya...>]

For those who are interested, here is a document that discuss the main parts of
the Embedlets Group's strategy.  Up until this point, it has been a somewhat
informal strategy but this is in the process of changing fairly quickly and the
following materials will probably be used as a starting point for developing an
official Embedlets road map.



WHY THE EMBEDLETS GROUP WAS FORMED

The reason that the Embedlets Group was initially formed was to discover 'the
most likely markets for Embedded Java' because, even though Embedded Java
systems like the TINI came out in 1999, Embedded Java was not having the amount
of market uptake that it should have. One of the more critical things we
eventually discovered is discussed the following 'Bridging Technology' section.

It is important to point out, however, that the Embedlets Group was designed to
be a place where Embedded Java vendors and Embedded Java open source developers
could work together to promote all facets of Embedded Java.  To this end, we
are definitely following a principle that Sun Microsystems continuously
stresses which is 'cooperate on specifications and compete on implementations'.

We are currently still in the 'specification creation' phase but we will soon
be bringing the 'compete on implementation' phase online and a vendor's area
will be added to the Embedlet's web site to accommodate this in the near
future.



EMBEDDED JAVA IS A BRIDGING TECHNOLOGY

The background information that explains why Embedded Java is currently a
bridging technology can be found on the Embedlets main web page:

http://embedlets.org


This diagram:

http://embedlets.org/images/flounder.jpg


And if that is not enough for you, in this OpenOffice presentation:

http://tkosan.javadevices.org/misc/embeddedjava/embedlets_jacknet_presentation.sxi


What this information states is that (at this point in time) the Embedded Java
community's main strength is that it is a bridging community which sits between
the traditional Embedded Systems community and the Enterprise/CS software
development community.

These two very different communities need to be able to communicate with each
other before the vision of 'billions of devices being brought to the internet'
can become a reality.  They are so different, however, that direct
communication between them is virtually impossible and so a translator
community is needed.

By definition, an Embedded Java developer has an intimate understanding of both
these communities so is in an excellent position to provide this needed
translation function.

As the flounder diagram shows (http://embedlets.org/images/flounder.jpg), we
are a very small but, potentially, a very important community.



THE EMBEDLETS GROUP'S STRATEGY

Once the above insights were realized, forming a multifaceted strategy for
putting Embedded Java 'on the map' was relatively straight-forward.  Here are
the main steps in this strategy:


1) Form a small group of Embedded Java developers that will then create the
open software and hardware foundational technologies and knowledge base needed
to bridge the traditional Embedded System's communities and the Enterprise/CS
software communities.  

Status: The Embedlets project was officially formed in December of 2002, the
Embedlets container code base was mostly functional by late Spring of 2003 and
we are about to figure out how JAPL devices should be integrated with
Embedlets.  The initial group was purposely kept small so that focus could be
maintained but we will soon be enlarging the group's membership (see #3 below).



2) Select an initial target market for this technology, create a class of
product that will be desired by this market and then develop a strategy for
approaching this market with the product class.  

Organizations that need to integrate their processes with their backend
Enterprise systems (like 6 Sigma organizations) was selected as the initial
target market.  The class of product we are in the process of developing is the
'Enterprise Outpost' and the unique strategy that was developed to approach
this market was to create an Enterprise Outpost reference implementation, write
white papers and articles about Enterprise Outposts targeted toward J2EE
developers about it and demonstrate Enterprise Outposts to J2EE developers at
the most promising venues (like the JavaOne conference).

Status: Using the criteria of quality, functionality, price and maturity, TStik
has been selected as the Embedded Java system to be used for the Enterprise
Outpost reference implementation and we are currently in the process of
figuring out a reasonable way to package it.  As soon as the packaging and
production issues are resolved, and step 3 below is put into action, we will
then move forward with creating the white papers, articles and demonstrations
needed to market the Enterprise Outpost concept.

<note>
Even though the Enterprise Outpost is still the most promising technology to
move forward with, it appears that during the next year, the work we are doing
with it can possibly be leveraged to create a Home Automation Outpost.
</note>


3) As soon as steps 1 and 2 have reached a reasonable level of maturity,
increase the size of the Embedlets group by starting to openly advertise our
existence in the most popular Embedded Java lists.  Also, encourage traditional
Embedded Systems developers to join the group by providing them with free
online educational materials which are designed to teach them about Java in
general and how Java can be used to great advantage in Embedded Systems.

Status: It is looking like the month of December will be a good time to start
officially advertising our existence in the main Embedded Java lists.  Since we
appear to be the only vendor-neutral open source Embedded Java community in
existence, there is a reasonable probability that we can recruit the lion's
share of this community for membership in the Embedlets group.

As for the educational materials for Embedded Java beginners, we have a free 7
week online Java beginner's course ready to add to the Embedlets web site along
with a 5 week TStik/TINI beginner's class that is a follow-on module to this. 
A muvium beginner's module is also currently in the works and hopefully people
will use the 7 week beginner's class as a core around which modules can be
built for any Embedded Java systems that people are interested in.

Last, but not least, a version of BASIC (called JBASIC) has been developed
which has been designed to run on most Embedded Java systems and make them
*much* easier to use by Java beginners.  Hopefully this will be a technology
that can be used to help bring the large BASICStamp, PICBasic, etc. communities
over to the Embedded Java platform (finally!).

The educational materials, along with a beta version of the JBasic environment,
will be ready to add to the Embedlets site by sometime in December.



4)  After step 3 has been put into play, start encouraging J2EE (and other
backend and CS style developers) to join the group so that we can help them to
start leveraging the Embedded space.


Ted

__________________________________
Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard
http://antispam.yahoo.com/whatsnewfree