Re: HFOSS: VizAudio & DOTS

[Willie Walker <William Walker Sun COM>]

Hi All:

Bryen - thanks for writing this up.  :-)

> http://mail.gnome.org/archives/gnome-accessibility-list/2009-February/msg00046.html
vs.
> http://mail.gnome.org/archives/gnome-accessibility-list/2009-February/msg00056.htm

I think both have strong positives and strong negatives.  It's hard.  I 
think the following considerations need to be made:

1) Is it a good HFOSS project?  By that, I mean is it something a 
relatively inexperienced individual could sink their teeth into and be 
proud of what they accomplished after 10 weeks?  Can they end up with a 
good deliverable at the end?  Can they learn something about a domain 
they might not have otherwise been exposed to (e.g., a11y, GNOME)?  Can 
it have an impact on the direction of their career?

2) Is it a good GNOME project?  Can we get useful code to use in GNOME? 
 Do we have the potential to attract another community member?  Will 
the code have longer term impact for GNOME or the larger a11y community? 
 In addition, even in the event it might be "throw away" code, can it 
spur interest in an area, can it cause activity to happen, etc.?

3) Is the project set up for success?  Is the task manageable?  Can it 
be done in 10 weeks?  Do we have strong mentors in place?

What I see is this:

* VizAudio has good impact in a space we need in GNOME.  As an aside, 
but not as important, it can also be a visually cool thing to show off. 
 It also gives an intern an opportunity to do a summary of the 
state-of-the-art, which is useful.  We have Bryen who can help with the 
UI design so we will end up with something that works for at least one 
user.  :-)  I don't see it succeeding without a strong mentor or a 
self-starting student with a strong hacker mentality, though.

* The Dots stuff fills a void, but I'm not sure a standalone GTK+ app is 
the optimal solution.  I've spoken with several people in this space and 
they tend to believe integration with OpenOffice would be more 
appropriate and would also have a larger impact across multiple 
platforms.  Going the OOo way may be more complex, however, and may not 
get done as a result.  So, I want to be careful about potentially 
squashing the GTK+ way just because it may not be the ideal thing to do. 
  If the standalone GTK+ app can do things like flush out liblouis 
bugs, give us a better understanding of liblouis' capabilities, and 
develop a model that might transfer to OOo, then it would be worthwhile. 
 We also currently have Eitan as a strong and proven mentor.  We'd 
need, however, someone with very strong braille skills to help design 
the thing.  I also wonder if 10 weeks is long enough.

So....hey two interns, three projects.  Maybe we let the interns decide 
what they want to work on?  If that's the path we take, let's still at 
least try to line up a technical mentor for the VizAudio task and a 
mentor with braille expertise for the Dots task.

Will