Open vincent-lg opened 6 years ago
Hello, Thank you for this great suggestion. I also have this idea since a while. I hope to implement it soon!
hi are you planning it for next ver?
Yes, I'll try to implement this for the next version. I'd like to release a stable version for early August.
Hi, I'd like to integrate this feature for January. @vincent-lg, @zstanecic: if a piece of text is in bold and underline for example, should we combine the landmarks like this? ⣇⠃⠥⣸blah-blahh-blah⣇⡃⡥⣸ Also, if we press a routing cursor button above '⣇⠃⠥⣸' or '⣇⡃⡥⣸', what behavior would you like?
In my case, I'd like the Cursor jumps to the nearest char it the landmark belongs to. What do others think?
Hi, Sorry answering took so long. I do agree with @bdorer. It would make sense if clicking on the landmark jumped to the first character in the landmark. NVDA behaves like this for other things, I believe, like HTML landmarks (ln, edit and so on). Hope that helps, Vincent
@vincent-lg, @bdorer Thank you for your feedback. I agree with you. I plan to resume development soon (this summer).
I propose that, instead of using dot 78 to mark a formatting change, Braille delimiters be used to clearly identify different formats.
Justification
Adding dot 7 and 8 to inform about a specific formatting offers several drawbacks:
As it is, the user relying only on Braille (as I do) would turn on only one of the options of Attributa to avoid clustering and hard-to-decode information. For instance, I like to see my spelling errors with dot 7 and 8, but if I turn anything else on, I won't know what's what.
Proposal
Instead of using dot 7 and 8 to indicate these formatting changes, I propose to use clear Braille landmarks that, maybe, the user can choose. There would be one Braille landmark before the affected portion, and another one at the end. For instance:
I know this will add quite a layer of complexity, since it will shift Braille cells further and render NVDA's calculation for the routing cursor invalid, but I believe this is worth it.
I have detailed this proposal on a NVDA issue but have been asked to talk to you about the problem first.