Here's an idea: someone should produce a series of ebooks that teach how to use GNOME and popular applications (e.g. Firefox and LibreOffice) with Orca, in a systematic way with activities for each concept or task, then use the proceeds from sales of those ebooks to help fund development of Orca and the free desktop accessibility stack in general, as NV Access has done with NVDA and their training ebooks. I think there would be a market for that.
Federico Mena Quintero reshared this.
Matt Campbell
in reply to Matt Campbell • • •TheEvilSkeleton đźđł đłïžâ§ïž reshared this.
Matt Campbell
in reply to Matt Campbell • • •Of course, GNOME and (especially?) Firefox are subject to UI churn as well, and Mozilla is now adding AI where people don't necessarily want it. But maybe it's not as bad (yet?) as what Microsoft has been doing with Windows 11.
One possible mitigation might be if the training material was aligned with a recent LTS distro. Maybe the recently released Debian 13 (trixie)?
Danielle Foré
in reply to Matt Campbell • • •Matt Campbell
in reply to Danielle ForĂ© • • •Danielle ForĂ©
in reply to Matt Campbell • • •Matt Campbell
in reply to Matt Campbell • • •Darrell Bowles
in reply to Matt Campbell • • •TheEvilSkeleton đźđł đłïžââ§ïž
in reply to Matt Campbell • • •GNOME Release Calendar |
GNOME Release Calendar |âïžkarolherbstâïž
in reply to Matt Campbell • • •though Red Hat _does_ backport features if there is a good reason of doing so, so maybe it makes sense to bring it up. But without a paying customer (well.. all state agencies here) pushing for it, there isn't a big incentive to do so.
However if it's deemed to be important, it might get backported by Red Hat engineers.
Adrian Vovk
in reply to Matt Campbell • • •Matt Campbell
in reply to Adrian Vovk • • •Matt Campbell
in reply to Matt Campbell • • •Correction: RHEL 10 backported the keyboard monitor implementation in Mutter, and updated at-spi2-core and Orca to versions with that feature. Thanks to @AdrianVovk for pointing out that this was likely.
Still, I think it might be better to align with Debian 13 (trixie), which ships GNOME 48, for my training idea, both because it has a slightly newer version of GNOME (and presumably other components too), and because the upstream distro is fully free.
Rene Ludwig
in reply to Matt Campbell • • •Darrell Bowles
in reply to Matt Campbell • • •Syntax
in reply to Matt Campbell • • •Matt Campbell
in reply to Syntax • • •Orca
OrcaSyntax
in reply to Matt Campbell • • •miki
in reply to Matt Campbell • • •and I'd say:
1. Pair this with a blind-friendly distro that has everything configured (you'd need NFB/ACB support for this, as you probably need the metaphorical "company letterhead" to get secure boot certified).
2. Autotranslate into all the languages via LLMs. Sure, human translations are better than automatic translations, but automatic translations are better than nothing. Source: Not a native English speaker, I've actually relied on them back when Google Translate was borderline unusable.
3. Preferrably, pair this with scripts that can "walk through" the scenarios described in the book, making sure Orca output stays consistent as versions change. You could also this to automatically record multilingual walkthroughs, with human-written commentary between the steps.
miki
in reply to miki • • •Matt Campbell
in reply to miki • • •I doubt we'd need a fully custom distro, with its own boot loader and unified kernel image that would need to be specially signed for Secure Boot. A stock boot loader and UKI from one of the major distros should be enough. For the rest of the distro, a Debian Pure Blend (debian.org/blends/) might be enough.
I really like the idea of using the training activities as regression tests.
Debian Pure Blends
www.debian.org