I think our current implementation of additional outputs (Braille, Mag, other Non-speech) is rather limited.  I don't know the inner workings of this perk but I suspect this bug is a good example.  For instance, in FirefoxPerk, we utilize the review perk knowing that Braille (and perhaps others) are catching these events.  This is fine, but I feel it does not always suit our needs or is awkward at best.  I think one solution would be to create some sort of 'universal' output task that each output type would have to support.  State information within the output devices may pose a problem, but the idea might be explored.

Any Perk can always use say/send to implement its own Braille, mag, speech interface. But to avoid duplicating a ton of code, we try to play off of what's in the basic perks as much as possible. Otherwise, Firefox (for example) has to remember to update the Braille display every time a review is performed as well as the magnifier. This gets ugly, very quickly.

lsr (Linux Screen reader) development has been stalled and it has been unmaintained for a few years now. Maintainers don't have future development plan so i am closing the bugs as WONTFIX. Please feel free to reopen the bugs in future if anyone takes the responsibility for active development.