Thread Subject: Re: Combined HardwareSubcommittee Proposal

Note

This archival content is maintained by WebAIM and NCDAE on behalf of TEITAC and the U.S. Access Board . Additional details on the updates to section 508 and section 255 can be found at the Access Board web site.

From: Peter Korn
Date: Tue, Dec 19 2006 4:20 PM

Hi Randy,

In the case of the Rock Box, where a user re-flashes the BIOS on an iPod
to produce a new - and accessible - user interface, you wrote:

> ...
>
> It brings up an interesting point â we usually think of assistive
> technology software as something that co-operates with the operating
> system to provide accessibility. Here is an example where something
> completely replaces the operating system, perhaps making the hardware
> more accessible. But does that mean Apple can claim the iPod is
> accessible? Certainly the Apple-produced software on an iPod still is
> not. The system is âopenâ because someone took a can opener and pried
> it open... (so to speak). In my mind, the package as a whole (OEM
> hardware /and /software) is still self-contained.
>
> ...
>
> So, back to 508. Perhaps our definition should include something like
> this:
>
> A device is considered âSelf-Containedâ if the /original manufacturer/
> has not provided a way for third-party software and hardware to be
> added to the device.

By this definition, I would say that Windows 3.1 and (frankly) Windows
95 are both "self-contained" systems. Screen reader vendors took,
essentially, a "can opener" and pried open the display sub-system to
observe the text calls to build their off-screen models to provide an
alternate user interface. The original manufacturer had not provided any
explicit way for screen readers to be hooked into the system. They used
hacks and reverse engineering and patching/replacing components of the
OS in order to do their job.

I do see the Rock Box case as different from the Windows 3.1/95 screen
reader case, but the proposed definition above doesn't distinguish those
cases for me. In one case (Windows 3.1/95 screen reader) vendors used
system calls to modify code in memory to patch the display driver - or
wrote to disk and replaced the display driver there. In the other case
(iPod/Rock Box), vendors used BIOS reflashing capabilities to replace
the entire software package with their own. The platform facilities were
used in both cases - just for purposes unintended (and unsupported) by
the original manufacturer.

The difference I see with Rock Box vs. Windows 3.1/95 screen readers is
that the AT isn't providing access to an existing vendors product.
Rather it is creating an entirely new product that only re-uses the
existing vendors hardware. Kinda like buying an HP Windows PC, and
replacing Windows with UNIX. Entirely new product at the end (which
might "make the HP PC accessible", but it certainly isn't making the
entirety of what was purchased accessible).

If the Rock Box process (or some other) were to involve reflashing only
a portion of the iPod's software in a way that added these speech
capabilities and left the rest of the functionalityi nplace, then the
situation would be completely analogous to Windows 3.1/95 screen
readers. The only difference then would be that we are used to thinking
of desktops as being "open" and iPods as being "closed".


Regards,

Peter Korn
Accessibility Architect,
Sun Microsystems, Inc.

WebAIM is an initiative of:
Center for Persons with Disabilities (CPD) Utah State University