Braille Extender not saving tables


Deborah Armstrong <debee@...>
 

Nvda 2019.1.1 current stable. Braille Extender 2019.07.10, also current stable.
This problem was reported on the list by Rob Hudson back in January.
 
Though the settings let you configure two Braille tables, a primary and a secondary, pressing OK does not save them. I've tried ENTER, I've tried space, I've tried going to another field and pressing Enter.
 
The secondary table does save but not the primary so when you try to switch tables you get the message:
"you must choose at least two tables for this feature. Please fill in the settings."
 
I've tried going to the Braille extender settings through preferences  and Settings, then Braille extender.  also finding the same dialog just under Preferences.
 
I don't know how you can save more than two tables though documentation implies you can. 
Why aren't the tables I specify saving? Is this a bug or a misunderstanding on how the settings interface works?
 
--Debee
 


David Csercsics
 

It is definitely a bug! I can duplicate it here as well as a couple other things. But the add-on will likely no longer work as of 2019.3 because of the move to Python 3 support, so we may have to find another solution.

 A pity for me because I really enjoy my braille display, but I'll see what I can learn to sort this out.


Luke Davis
 

Why do you say it may no longer work because of Python 3 support being lacking?
According to its github page:

https://github.com/Andre9642/BrailleExtender

It seems to be under quite active development, with multiple developers, and claims compatibility with NVDA 2019.4, which at least implies that it is already working with Python 3.

It doesn't hurt to write to the developers and report these bugs.

Luke

On Sun, 28 Jul 2019, David Csercsics wrote:

It is definitely a bug! I can duplicate it here as well as a couple other things. But the add-on will likely no longer work as of 2019.3 because of the move to Python 3 support, so we may have to find another solution.

 A pity for me because I really enjoy my braille display, but I'll see what I can learn to sort this out.





David Csercsics
 

I wrote to the developer, but I didn't check the git hub page, I was just saying it may not work with  Python 3 because I wasn't sure if he'd have time to work on it. But we'll see what we can do. I will ping him and see if he needs documentation help and possibly brush up on my python and help that way as well, since I have the time. I just haven't learned all the Windows API stuff I might need. But I'll see if I can get a build environment rigged here and see what can be learned from that.


 

Hi,
Just because an add-on declares support for 2019.3 or later does not
necessarily mean it is Python 3 ready. Braille Extender, although the
manifest says it is ready for future NVDA releases, source code says
otherwise. For instance, the add-on won't load because a module located
inside the add-on isn't loaded properly because Python 3 can't find it.
Cheers,
Joseph

-----Original Message-----
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Luke Davis
Sent: Monday, July 29, 2019 10:47 AM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Braille Extender not saving tables

Why do you say it may no longer work because of Python 3 support being
lacking?
According to its github page:

https://github.com/Andre9642/BrailleExtender

It seems to be under quite active development, with multiple developers, and
claims compatibility with NVDA 2019.4, which at least implies that it is
already working with Python 3.

It doesn't hurt to write to the developers and report these bugs.

Luke

On Sun, 28 Jul 2019, David Csercsics wrote:

It is definitely a bug! I can duplicate it here as well as a couple
other things. But the add-on will likely no longer work as of 2019.3
because of the move to Python 3 support, so we may have to find another
solution.

 A pity for me because I really enjoy my braille display, but I'll see
what I can learn to sort this out.






David Csercsics
 

Yes, that's what I saw, too, when I briefly looked at it. It requires syntax changes and some testing, and the author may not have time to do it. How much of the functionality should we put in NVDA core, possibly? Having to install an add-on for extra braille features seems a bit weird. But then again, I understand about keeping the main screen reader as portable as possible. This is probably for another thread, but I was thinking about it because NVDA commonly gets blamed for poor braille support, because people don't know about the Braille Extender add-on. It looks like a lot of the functions are just conveniences, but maybe some should be part of NVDA, I'm not sure. Maybe this belongs in another place than here, but since we're on the topic of refactoring it seemed related. I don't know how many braille users we have either, so maybe the community doesn't need it that much.


Deborah Armstrong <debee@...>
 

Guys, this was my mistake! It works perfectly. The problem was the help screen, you are supposed to read by pressing F1,  wasn't reading on my home PC. I tried it at work and could read the help fine.
Here's what you do in Braille tables. You arrow through the list of tables and press Spacebar on the one or ones you want. The first time you press Space it says "input only" and the second press of spacebar it says "output only". The third time you press space it says "both input and output"  and the fourth press of Space cycles back to "none". By default, they are all set to "None".

It's not a bug exactly, just a documentation bug. It should be explained clearly in Braille extender's documentation how configuring tables works, and it's not.  Instead the documentation has details on the default keystrokes for Braille extender which you can find under input gestures anyway.

I think I'm going to post something over in the chat subgroup about documentation bugs vs real bugs, it's an interesting topic, but not relevant here on the main list.


David Csercsics
 

Yes, that makes sense. He did say somewhere that the documentation needs to be improved. We could see if that’s possible, assuming the add-on will get updates once they move to Python 3.