Thursday July 22, 2004 
The collected occasional commentary by Peter Korn, Accessibility Architect at Sun Microsystems, Inc.
|
|
Peter Korn's Weblog The collected occasional commentary by Peter Korn, Accessibility Architect at Sun Microsystems, Inc. |
Thursday July 22, 2004 GUADEC '04 Trip report (aka Accessibility@GUADEC)GUADEC in Kristiansand was my second GUADEC (first was the one in Dublin). Like GUADEC '03, (or should I say "GU4DEC"?), there was lots of accessibility content this year, and I counted four blind individuals among those who made the trek to Norway.
The overall sense of the BOF was getting stuff out to users in a usable state, and into distros so people could use it. We talked about the name issue: "GNOME Accessibility" isn't so appropriate when the new KDE/Qt Accessibility work is sharing much of the same architecture. While the KDE folks in the room didn't think this was a barrier to existing engineers, going to a more neutral name was seen as "immensely helpful", with "Free Desktop Accessibility architecture" as a suggested candidate. Janina said she would assemble a list of all GTK+ 2 based music apps for accessibility testing/use (a personal interest of hers). Much of the rest of the discussion was around a "to do" list of things remaining.
Accessibility "to-do" list:
Bill then talked about some of the Really New Stuff: the COMPOSITE extension and it's implications for magnification; the Sphinx-4 open source speech recognition project; KDE's adoption of the GNOME ATK interface for accessibility; and the neat things we can do with Python and the gnome-python module - leading to the orca screen reader scripting playground. Next he observed some of the important accessibility lessons learned in the GNOME 2.6 project: we need much better regression control - there were Gtk+ issues with late integration of new widgets; we need to figure out how to detect and handle behavior changes; we need better Accessibility testing of new apps; and we must deal with the myriad configuration problems. The accessibility team needs help from the entire GNOME community. Bill discussed what is needed to make it work. Docs can be found at http://developer.gnome.org/projects/gap/guide/gad/. There is a testing article from the folks at Wipro, and an accessibility sanity test suite developed by Sun (details here). Of course there are also the test tools at-poke, brlmonitor, event-listener-test, and test-speech. Bill finished his prepared remarks by noting what is needed to "get it out there": support in GNU/Linux distros, a self-install for accessibility, and better help for users. Bill ended the presentation with a quick demo of Gnopernicus with BrlMonitor, speech, and magnification. He then invited Marc Mulcahy to the stage. Marc talked about his work over the past few months exploring screen reader scripting, and the orca playground/exploration project using AT-SPI from Python. Marc launched orca, and then proceeded to read www.sun.com via Mozilla (thanks to a custom orca script for Mozilla). Orca read the text in three different voices: a high-pitched female voice for hyperlinks, a low-pitched male voice for images; and a moderate male voice for the web page text. He then brought up the Metacity window-switch dialog, showing the two different voices that are used here to help the blind user understand what is going on (another orca script). Finally Marc launched the GAIM instant-messenger client, and people in the audience immediately got on-line into the same chat room and started chatting with him, which showed off the orca script which automatically spoke chat messages in the output window (so audience members' postings were broadcast to the room in speech). This full session ended on that high point, having managed to somehow squeeze into 60 minutes!
For Universal Access, he stated the goal for 2.8 is accessible login, and that the focus is on making things "just work" out of the box. For Collaboration, the work in 2.8 is on the Evolution Data Server, Addressbook & Calendar integration, and inclusion of mDNS shares; longer term work on indexing/search, metadata, and peer-to-peer sharing. In the Media facet, he noted that sound & video are important, but there are legal issues to work through; in 2.8 the work is on CD importing & burning, portable and audio devices; longer term work on the audio server, and media widgets in the platform. In Hardware, he noted the wide range of devices GNOME runs on , and the need for cooperation w/other communities & Project Utopia. Hardware goals in 2.8 are printer management, the volume manager, computer integration, and D-Bus/HAL; future work is on a wider range of devices (music players, bluetooth), and increased support for hardware in other apps (gPhoto, Rythymbox, etc.). For Manageability in 2.8, David cited the new panel menu system based on freedesktop.org standards, and easier-to-administer file associations; longer term work is a networked GConf backend, and decreased complexity of GConf for administrators. Finally on the Development Platform - defined as building the desktop, 3rd party developers/ISVs, and freedesktop.org - David cited the 2.8 goal of pango rendering in libgnomeprint; and the GTK+ 2.6 goal of migration of libgnome API into GTK+; with the longer term goals of a Cairo-based rendering API and printing API in GTK+. David's asked everyone to work to stay up to date: cooperate with the contributors; contributors should keep their eye on roadmap; the release team must poll the contributors before development cycles; and we all must help keep the roadmap up to date. After David's formal presentation was over, the entire GNOME Foundation Board (that was available at this time slot) came up on stage and answered questions. I didn't manage to capture many of the questions in my notes, but for me one of the most interesting things discussed was the idea of a fund to help facilitate the goals of the GNOME community - which individuals and corporations would be encouraged to contribute to. Small (e.g. $5,000) grants would be given for various things - examples cited included helping start a GNOME conference, or purchasing hardware devices needed for testing and development of accessibility. Explicitly NOT fundable is software development, which is seen as a community policy issue that the GNOME Foundation Board is specifically chartered to not address. I've felt for a while that we needed to look at alternative funding models for accessibility and assistive technology (more on that in the upcoming Libre Software Meeting trip report). While not addressing the larger question of funding ongoing assistive technology development, being able to use these funds to acquire needed hardware for testing and development would be a good development! (2004-07-22 15:02:24.0) Permalink |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
