Steve Lee <st...@fullmeasure.co.uk>

Hi, well everything has been quiet since I completed the last phase so
thanks for hanging in there and staying interested. I've had a bit of
a break due to personal circumstances and perhaps slightly running out
of steam. A big thanks to Eitan for all the nags during this time
(your mentoring has been great).

This is the time start thinking about the Roadmap and I would
appreciate contributions. I'll try to keep this brief in the hope you
stay with me and get inspired to have your say at this critical point.

So where are we? Well no one seems quite clear on exactly what Jambu
is and that was my fault for being too big and woolly at the start. I
had big ideals but not so much focus. What we have at this time is:

1) Idealistic ideas about improving all Alt input access
2) In application selection and operation of accessible GTK apps and
Firefox on Linux with simple switch access.
3) A GTK app or custom widget and framework that renders SVG UIs like
OSKs described in declarative XML-based representation.

So that is really 3 projects.

1) Is really a Meta project and I think too much to chew on at once.
It covers basic input devices/drivers, desktop accessibility features
like mousekeys/mousetweeks, a11y APIs, ATs, applications, DOM input,
WCAG etc.

Another consideration for alt input is that after seeing and using eye
gaze, it seems that this should become the default alt access mode,
with switch being used for those who don't get on with gaze. We need a
FOSS version that works really well with cheap-o web cams and
OpenGazer seems to be running in that direction. We can expect a
period of inertia with users getting out of the switch mind set, but I
think it will happen. So this encourages me to work in other areas.

2) Is unique and addresses those who want to use normal applications
without intermediate OSKs (this may be a small user base as we're
being speculative). It's rather cool and I think it should be
developed more.  Intitial work will be ensuring it works with recent
changes in FF3 and GAIL regressed a bit. Finding some users is
paramount in my mind. It may have other non a11y uses as well for low
bandwidth input situations.

3) my vision for Jambu has matured to be a tool for achieving
communication and control with SVG symbols, simple custom OSKs and
possibly multi-media.  This is important for both people with physical
and learning difficulties (or both), all of who are currently under
represented (note to self of a danger of getting too broad again). I
have been seeing a lot of discussion about these needs in education
and from Speech and Language therapists. I'm also working on a project
that will be an open set of symbols and there is synergy there as well
as a chance for more user contact. Several of the OATSoft team are
highly experienced in this area which is high profile in the UK in my
experience. There is also a semantic web based project called CCF from
Dundee University that aims to make translation to/from symbols
workable. So I see a need and have many contacts with people working
in this area.

There are several well established proprietary programs in this area
as well as an overlap with OSKs so I'm keen to ensure we are not just
a 'me too' OSS version but really innovate and enable participation. A
key point is it should be easy to create new specific solutions or
modify existing ones.  Rather like the way the Web provides free
access to the source and anyone has easy entry to creating resources
and solutions (however messily). We will need a user base for
community and creating a library of custom shared resources.

I'm seriously thinking how this should become a web based solution /
service usable standalone or online or as part of another application.
It seems clear that IT (and browsers) are moving in that direction
anyway and even if it is a while off perhaps a local webserver would
be easily installed for offline use. Another attractive possibility is
to revisit XUL now it's status and my understanding have moved on. I
won't go into all the thinking now but, really do want to thrash this
idea out more.

So if we drop 1) out of Jambu we still have to decide whether to split
out 2 & 3 so each is clearly their own project. I think this is
probably the way to go, perhaps making it clear they are siblings.

Finally an exciting possibility is getting Project Possibility
students involved and I'm keen to do that. I really need others
working on it now and later phases and student energy can be awesome.
There should be scope for subprojects in 1 such as porting SUNFlare
gesture recognition or filling alt access gaps at the Desktop level.
These will involve working with existing projects so offer good
experience. 2 will be excellent for students who are happy to take a
starting point and run with it. It gives you some python and AT-SPI in
your portfolio. 3) will also offer lots of possibilities once it is
more clearly defined, I see the possibility for several sub-projects.

So my immediate next steps are
* Create a proper website etc. so Jambu is easily approachable (honest
I will Eitan)
* Develop the Roadmap and then try to get funding

Looking forward to any thoughts ideas.

Cheers

--
Steve Lee
--
Open Source Assistive Technology Software
web: fullmeasure.co.uk
blog: eduspaces.net/stevelee/weblog