AxAp News, January 2017
Welcome to AxAp News, an online newsletter about the AxAp Project
This inaugural issue covers our initial efforts, starting with the project's prehistory
and leading to the current, proof of concept prototype
It also discusses some of our plans for further development.
In early 2016, Rich Morin
started thinking about the possibility
of using standards and technologies from the World Wide Web
to improve the accessibility
of a wide range of digital content
Standards such as HTML5
encourage the use of accessible content on the web.
Using them for other content (e.g., data files, documents, source code)
seemed like a plausible and worthwhile objective.
Talking with Amanda Lacy
, Rich soon realized that the state
of accessibility on the web was far from perfect.
Although various standards exist, each user interface (UI
(e.g., operating system
, web browser
, screen reader
has its own way of implementing them.
More critically, most web sites ignore these standards now
and will do so for the foreseeable future.
Clearly, some kind of mediation software was needed.
Accepting various forms of content as input,
it would add accessibility markup (e.g., HTML tags, ARIA attributes)
and hand off the results to existing UI stacks.
However, Rich found significant problems with all
of the "obvious" approaches he considered
(e.g., modifying screen readers, writing browser plugins).
Modifying existing software presents a number of challenges.
Although some of the relevant software is open source
many popular programs (e.g., JAWS, Safari, VoiceOver) are not.
So, their internals can't be examined, let alone modified.
Even when the code is modifiable in theory, it may not be in practice.
Issues such as language choice, packaging, and distribution
can easily become overwhelming.
There was also the problem of market fragmentation
No single screen reader works on every operating system,
each browser has its own API
(etc) for plugins, etc.
Supporting this range of software (and its combinatorics
was simply a non-starter, especially for a small, volunteer project.
So, Rich needed a way to create a single, widely applicable solution.
More critically, browser plugins operate in a very constrained environment.
(or compiled into it).
They are also prevented (for reasons of security) from accessing various resources
(e.g., local files, I/O devices, and processes; networking).
So, many of the features Rich had in mind would be impossible for a plugin to support.
Rich's response to these challenges was AxAp, an "accessibility application".
It would be constructed as a personal web server
(of sorts :-),
running with all of the privileges of normal user process.
It could run on any operating system, access any needed resources,
and use infrastructure written in any language.
After a few months of requirements analysis
and systems design
Rich started work on
(AxAp Demo Server),
a testbed and proof of concept
implementation based on Ruby
and the Sinatra web framework
About 7000 lines of code later,
is now giving Amanda a more convenient way
to read both EPUB
documents and program source code
as well as making a start on presenting data files
This wiki page is maintained by Rich Morin
an independent consultant specializing in software design, development, and documentation.
Please feel free to email
comments, inquiries, suggestions, etc!