Re: [orca-list] Bug in speechdispatcherfactory.py

[Alexander Epaneshnikov <aarnaarn2 gmail com>]

08.01.2021 14:58, Joanmarie Diggs пишет:

> Hey Storm.
>
> Sorry for the delay in responding. Playing catch-up after holidays.
>
> As I indicated in the commit I just made to Orca master: Those lines
> were commented out years and years ago. When I re-enabled them, other
> stuff broke, so I commented them back out.
>
> I'd be interested in knowing what specifically you are seeing in
> keyboard interrupt not working. While the lines in question are indeed
> commented out, the default script's presentationInterrupt method is
> what should be called for keyboard interrupt. That method interrupts
> both speech (by calling speech.stop, which in turn calls _cancel) and
> braille (i.e. clearing any flash messages).
>
> Having said that, I do think it's silly to have perpetually commented-
> out code preventing an optional argument from doing anything. :) In
> light of all this, the change to master I committed uncomments out the
> lines in question, but also makes the default value of interrupt False.
> This should minimize the instances where we are now interrupting speech
> where before we weren't. Maybe. I hope. ;)
>
> Does this change fix whatever "doesn't always work like it is supposed
> to issue"? Mind giving me some steps to reproduce the problem you're
> seeing, regardless?
>     
    
    hay Joanie commit ce033c086
      introduces a big regression for me. 
      steps to reproduce:
      1. update orca
      2. open preferences
      3. use tab or arrows to navigate
      what happens:
      after a couple of seconds orca says only 1 out of 6 2 out of 6 etc
      ...
      the only way to read the contents of a window is to use flat
      review.
      
      I did not study the behavior further, but simply reverted this
      commit.
    
> Thanks!
> --joanie
>
>     
    --
Sincerely, Alexander.