Regression in NVDA2024.1: brltty cannot connect to Metec40 device
Closed this issue · 13 comments
Could this be related to #15658?
cc @LeonarddeR
Steps to reproduce:
- Install the brltty application.
- Connect the Metec40 to the computer.
- Select "brltty" in the NVDA Braille Device dialog
Actual behavior:
NVDA throws error:
snippets_log.txt
Expected behavior:
Metec devices should connect and work as expected, just like in nvda2023.3.4.
NVDA logs, crash dumps and other attachments:
System configuration
NVDA installed/portable/running from source:
Portable, Installed
NVDA version:
2024.1, 2024.2Beta3
Windows version:
Windows10
Name and version of other software in use when reproducing the issue:
Brltty version is currently unknown
Other information about your system:
None
Other questions
Does the issue still occur after restarting your computer?
Yes
Have you tried any other versions of NVDA? If so, please report their behaviors.
2023.3.4 and earlier versions work as expected
If NVDA add-ons are disabled, is your problem still occurring?
Yes
Does the issue still occur after you run the COM Registration Fixing Tool in NVDA's tools menu?
Yes
This could theoretically be because of the switch to Python 3.11, but I think it is very hard to debug this.
If you need any help debugging please let me know.
I will also check further to see if it is related to the version of brltty and the Windows.
See also #16566. I do not know if the cause is the same.
Looks like #16566 is very similar to this.
For Metec, NVDA can almost only use Brltty to establish connections. If this problem is not solved, it means that these devices will completely lose their value.
Could you please report this to brltty as well?
https://brltty.app/contact.html
Could you please report this to brltty as well?
I hope to do this after getting more information about BRLTTY.
@cary-rowen - could you review bramd's message here: #16566 (comment)
@LeonarddeR If I remember correctly, you helped compile the Python bindings for BRLTTY when we upgraded to Python 3.11. Would it be worthwhile to recompile with the specific minor Python version we use now and if so, do you still have a dev environment for this set up? It's not that easy to get the whole BRLTTY compilation setup going on Windows.
I'm also afraid this might be hard to debug. I'm not really familiar with the Python bindings, but the error about the invalid file handle might indicate that the handle is somehow closed/gone or the BRLTTY process died somehow making the pipe invalid. If so, a log from the BRLTTY side would also be nice to see here.
@bramd One of the things I did while compiling is providing an appveyor script to do it. So recompiling it shouldn't be too difficult. Having logs from the brltty side of things could definitely help, though.
It looks like MinGW is missing on Appveyor now. I filed appveyor/ci#3925
I have been unable to reproduce this using brltty-win-6.6-0-libusb-1.0.