Creating An Accessibility Specification for Lubuntu 11.10

Jonathan Marsden jmarsden at fastmail.fm
Fri May 27 09:07:13 UTC 2011


Pia (and accessibility team),

On 05/26/2011 03:22 PM, Pia wrote:

> What John is asking for seems so obvious to us who are disabled that
> I forget normal people don't "get it".

I really hope the Ubuntu Accessibility team is not composed entirely of
folks who are "disabled" -- some more or less "normal" people may well
have an interest in accessibility issues, too.  Is dividing human beings
into "us" and "normal people" really a helpful and appropriate mindset
for an accessibility team?  All concerned might benefit more from
working together, than from creating artificial and unhelpful divisions
between people.

For a little perspective: I am the guy who spoke up at a UDS session
discussing possible software features for Lubuntu this release cycle and
said (via IRC)

  "do we need to consider improving accessibility features of Lubuntu
   in Oneiric ?"

(That is a direct quote, copied from my IRC log of 10 May 2011).

Had I not done that, as far as I know doing this work would not even be
under discussion for this development cycle, because accessibility would
not be in the blueprint for Lubuntu 11.10 at all.

I am an advocate for doing some work in this area in Lubuntu; in order
for a very small team to do that effectively and efficiently, we need
clarity on what exactly that work *is* , and how to prioritize it.  This
is not optional, it is required, if useful work in this area is to
happen in the next few months within Lubuntu.

Open source software development in general now has a fairly well
established set of stages, and Debian and Ubuntu software development
has its own perhaps even more specific variant of those.  Defining
clearly what it is you are trying to do, that is, writing a
specification, or blueprint, is an early part of that process.
Launchpad supports this with what it calls blueprints, UDS is the usual
venue for refining these specifications, etc.  There are probably books
and academic papers written about this process...

I submit that your stating or implying that I am "normal" and that I
"don't get it" were both rather unfortunate and unhelpful choices to make.

> That kind of request is like demanding that there be a specification 
> first that monitors and video cards have to be supported or that we 
> need mouse support.

(It is tempting to make a comment about non-software-developers not
"getting it", but that might be unkind).

Allow me to use your own examples: Does Lubuntu 11.10 need to add full
support for Tektronix vector graphics terminals as a primary output
device, or do we need to add drivers for Hercules monochrome graphics
cards this cycle?  Do we need to support the Xerox Star mouse?  Should
we go to special lengths to add support for one-button mice, since Apple
makes those?  Is testing that Lubuntu works well with Microsoft bus mice
appropriate and necessary, since no-one tested that in the last (Natty)
development cycle?

Saying "we need monitor, mouse and video card support" is a hopelessly
vague specification, in need of much refinement before it can be
implemented by software developers.  So is "we need accessibility"!

If you do not understand the rationale for software specifications being
created and refined before software is designed and coded, it will be
very difficult for you to make a significant contribution to the
software development process of Lubuntu.  I would therefore urge you and
the accessibility team to make a little effort in that direction, just
as I am trying to make some effort to learn more about what it will
really take, and what it really means, to "add accessibility" to Lubuntu
in a useful way.

We can help each other, if we choose to do so.  The alternative is to
declare that those who are not like us and do not share our own
background and education "don't get it".  Which would you prefer?

> The road map document would be simple.  We need:

I think this is a potentially useful initial rough draft; please do
publish it on a wiki page, and so allow it to be edited and revised by
the accessibility team as a whole, and to have others outside it
(Lubuntu developers included, if they wish) add comments and questions.

Here are a few questions from my initial reading of your list:

1. Are these items in priority order?  If not, can your team please
   order them, so we can consider implementing the most important ones
   first?

2. How many (and which) of them do Xubuntu, Kubuntu and Ubuntu already
   fully support? (This helps us understand how far behind Lubuntu
   currently is).

3. How many and which of them do Xubuntu, Kubuntu and Ubuntu plan to
   work on implementing this (11.10 Oneiric) development cycle? (This
   helps us collaborate and use any such current work, avoiding any
   unnecessary duplication of effort).

4. "Something that allows for braille output" might be more of a
   hardware specification than a software one.  I know braille output
   devices exist, and back in the days of text mode computer use have
   helped set them up.  Is there a class of these devices that already
   have working driver support in the Linux/Debian/Ubuntu world?  How
   do Ubuntu, Kubuntu and Xubuntu support them currently?

> special keyboard layouts and settings for the mobility impaired that
> will allow them to use toggle keys since they may not be able to hold
> down more than one key at once
> 
> Something to modify mouse behavior for the mobility impaired.

5. These seems to assume all users with mobility impairment will be
   able to use some form of keyboard and mouse -- is that realistic?
   I have seen "suck/blow" tube interfaces to a screen in (old) video
   about severely mobility impaired (mostly paralyzed) disabled users
   using computing technology; such users might find "special keyboard
   layouts" and "something to modify mouse behaviour" insufficient.  I
   have no idea if that is a common issue, or so unusual it can safely
   be ignored for Lubuntu.  Can it?

> A dictation package that responds to voice commands as the interface for
> the mobility impaired.

6. Are there any such open source dictation tools currently in
   existence?  Can you provide pointers to information about them?
   Is their memory footprint low enough for inclusion in Lubuntu?
   We definitely would not have enough developer resources to create
   such a product from scratch in 11.10.

> meaningful icons for the dyslexic but alt tags or meaningful text on the
> icons for the visually impaired.

7. I'm not sure I understand this item yet; can you provide examples of
   "good" and "bad" practice in this regard in the current Lubuntu
   11.04 default user interface, to help clarify this for us?

> See, list done.  Most of this can be done just with settings in the
> existing desktop, others require programs.

8. Really!  Can you (or your team) annotate your list with which items
can be done just with settings in the existing (Lubuntu 11.04 LXDE)
desktop, please?  And please could you also mention (and provide links
to web sites for) any specific programs you have in mind for all those
list items requiring programs?

> Mentioning espeak was important because the solutions also have to be
> lean and so it doesn't make sense to evaluate heavier resource
> intensive apps.

It doesn't make sense to evaluate *any* apps at all without a clear
definition of what you are trying to accomplish by adding those apps to
Lubuntu; hence my request for a clearer specification.

Thanks for helping us take the first small steps in this direction,

Jonathan

[ Throwaway aside of the day: why does a team labeled "accessibility"
choose to use a closed (i.e. inaccessible) mailing list?  On the
surface, that appears paradoxical.  Any chance you could at least make
the ubuntu-accessibility mailing list archives more readily "accessible"
(public) to the rest of us, so we can learn from reading them?  Ideally,
please make your list as open as this one (lubuntu-desktop) is. ]



More information about the Ubuntu-accessibility mailing list