What actions to enhance a11y supprt at LOdev 4.0?

[Michael Meeks]

Hi Stuart,

On Sun, 2012-11-11 at 12:37 -0800, V Stuart Foote wrote:
> I'll throw this out for the Dev community to mull over.

	:-)

> The apparent ESC decision to accelerate adoption of LibO 4.0 and not release
> LibO 3.7 would seem to eliminate a potential logical break point for
> implementation of several outstanding /a11y/ accessibility changes.

	Break for implementation ? it's already implemented AFAIK - but is
still not made available under the ALv2 (sadly).

>  Notably getting the Windows client off of the Java Access Bridge and onto
> a "native" bridge implementing IAccessible2 API.

	A useful goal - one we've had for years. As/when Apache get their code
into some usable state we can include it; currently it looks like this:

https://svn.apache.org/repos/asf/incubator/ooo/symphony/trunk/main/winaccessibility/source/UAccCOM/AccAction.cpp

> Apache OpenOffice will roll out their 4.0 release with IAccesible2 support
> from the /winaccessibility IBM Symphony contribution.

	As/when it's available with clear and clean licensing we'll include it
of course :-)

> Has there been any effort yet on migrating under ALv2 the /winaccessibility
> contribution from Apache svn to git for  a git build?

	As above really. The reality is AOO graduated with many millions of
lines of LGPLv3-only + IBM proprietary / confidential + copyright
headers in them, many of which are still there sadly.

> And while undertaking integration of a native IAccessibile2 for windows,
> would seem to be an excellent time to rework the UNO Accessibility modules
> to add additional roles and more completely implement outstanding features
> from the /a11y/ ATK/AT-SPI specifications and possibly exposing roles to
> Apple NSAccessibility or even Microsoft UIA.

	Really - there is nothing stopping us doing any of this at all, except
for the notion that it is time wasted if it has already been done -
we're waiting only for IBM to actually contribute it properly. I'm sure
that's frustrating for all impaired users on Windows - I know it's
frustrating for me and I imagine you.

	Anyhow - as soon as it's available; it is (I hope) easy to integrate -
and there are several ways we can do that; it can be done for 4.1 which
would be 6 months later - or, if we get a tripple review from 3x
developers we can put any feature into even the stable branch - so
potentially a 4.0.x release - though I'd (personally) not give that
review unless it was working well in 4.1 builds.

	I'm sorry for the lameness - but ... it's not a new problem - it's been
going on for years now.

	It's worth noticing that the winaccessibility/ code itself is almost
exactly 10k ';' lines of code - (including the imported IAcc2 IDL) - so
the bridge itself is pretty small / simple.

	ATB,

		Michael.

-- 
michael.meeks at suse.com  <><, Pseudo Engineer, itinerant idiot