Re: [orca-list] Small bug when using BRLTTY in a graphical terminal

[Sébastien Hinderer <Sebastien Hinderer ens-lyon org>]

Many thanks for your in-depth response Samuel, and sorry for the brevity
of mine.

I follow you and share most of your point.

Let me just add one thing one may want to take into account, even if it
can be considered to be slightly a side-note.

You were discussing how confident a screen-reader may be about its
ability to do some rendering. Although it will probably not change
anything at the protocol level, one thing to keep in mind, IMO, is that
this ability to render can also be seen as a user preference, i.e.
something that can change from one user to another for the two same
screen readers.

To give you a concrete example, take the Windows case. There, those
users used to NVDA may prefer its rendering of terminals, even if they
use BRLTTY to control their braille display. A user with a Unix
backgroound coming to use Windows will likely prefer BRLTTY's rendering
over NVDA's one.

Too bad the BrlAPI protocol does not have any way to let clients somehow
identify themselves. IF we had that, they we could imagine to configure
the priorities at the BrlAPI server level, which we can not do today,
unfortunately.

Sébastien.