View Issue Details

IDProjectCategoryView StatusLast Update
0006304Kali Linux[All Projects] General Bugpublic2020-12-01 10:50
Reporterzersiax Assigned To 
PrioritynormalSeverityminorReproducibilityalways
Status newResolutionopen 
Product Version 
Target VersionFixed in Version 
Summary0006304: Accessibility features appear to have seriously regressed since XFCE adoption
DescriptionPlease note that testing has, as of now, only been performed on a virtual machine, this due to the severity of the issues discovered.

I last assessed Kali's accessibility about 1.5 years ago, and was delighted to see that the normal interaction patterns can be used to activate Orca for blind computer users. This usually involves hitting ctrl+super+s on the login screen.

Unfortunately, that seems to have disappeared completely in the version I tried today, which is, I believe, based off of 2019.4 and likely still present in 2020.1.

My Findings:

- Orca seems to no longer be installed whatsoever, requiring a user to install it from the apt repos. On a VM, this is somewhat doable using OCR techniques in the host screen reader. On bare metal, the adventure just ends right here. This basically means that anyone who wants to use Kali on a dedicated machine, can't, right now. I have not tested this on a Raspberry pi yet but expect to find a similar situation there.
- After some troubleshooting, it appears PulseAudio might also not be enabled by default, making the screenreader unable to output audio, even if it was on.
- Even with both things fixed, I am unable to detect audio from the VM, which either means the audio is muted, at a too low to register volume or in some other way not responding.
- It appears Orca, once installed, can no longer be triggered from the login screen , or, indeed, any screen apart from manually invoking it from terminal or the application finder, which is certainly doable but far from ideal.


I hope I don't need to explain just how much of a dealbreaker this is. Kali can easily be called the De Facto standard for pentesting distros and any tutorial, course or training will expect you to at least be able to start it up and log in. As it stands, Kali seems to be unable to be used by security enthusiasts and would-be pentesters, analiysts etc. if they require a screen reader, such as myself.
I would therefore really appreciate if this can be looked into. I am a web developer / accessibility expert who happens to be blind, and I am working on a project to catalog and map out a set of tools and resources for other screen reader users to make their way into security. As it stands, I cannot add Kali to that list, and I really want to be able to. Please prove me wrong.
Steps To Reproduce1. To accurately reproduce the test case , create a Vmware VM using kali 2019.4 or 2020.1 as the image and install it.
2. Try ctrl+win+s on the login screen. In my case, nothing happened.
3. Jump into a text console and run apt-get install orca, in my case apt gave me output indicating this is indeed not currently installed.
4. Install Orca and run it, in my case no audio was heard.
5. This is where things get murky. Someone else working with me on this was able to get audio by running "systemctl --user enable pulseaudio && systemctl --user start pulseaudio" and rebooting the machine, although the audio was at an inaudible volume for them.
On trying this, I have not been able to get audio working, although I may just be unable to raise the volume as I am unaware of a procedure to do this from the keyboard.
This is where i got stuck.

Activities

zersiax

2020-04-28 10:32

reporter   ~0012687

Bump.
Some extra information where this is concerned after doing some research.

I dare offer that this issue only applies when XFCE is selected for the desktop environment. I am currently uncertain if:
- Orca is installed along with Gnome if Gnome is selected,
- What the state of pulseaudio is after an installation on bare metal
- What the expected strategy is for people requiring a screen reader who want to install Kali.

Given almost the entire installation procedure happens before actually picking a desktop environment, one has to wonder how this issue can best be mittigated.
Making sure speakup_soft is added to /etc/modules, and the espeakup package is installed on the live system would at least make sure a user can work through the text-mode install procedure ...if one manages to select that option at boot, that is.
Some thoughts would be very welcome here.

sec4lib

2020-08-23 01:44

reporter   ~0013311

Re: (...) " I have not tested this on a Raspberry pi yet but expect to find a similar situation there. " (...)

Indeed. You will have to install it from the apt repos. ( Raspberry Pi 4 Model B Rev 1.2 VERSION_ID="2020.3" "kali-rolling")

┌──(kali㉿kali)-[~]
└─$ sudo apt-get install orca
[sudo] password for kali:
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
  espeak-ng-data gir1.2-atspi-2.0 libao-common libao4 libatk-adaptor libbrlapi0.7 libdotconf0 libespeak-ng1 liblouis-data liblouis20 libpcaudio0 libsonic0 libspeechd2 python3-brlapi python3-louis python3-pyatspi
  python3-speechd sound-icons speech-dispatcher speech-dispatcher-audio-plugins speech-dispatcher-espeak-ng xbrlapi
Suggested packages:
  libsndio6.1 libttspico-utils espeak mbrola speech-dispatcher-doc-cs speech-dispatcher-festival speech-dispatcher-cicero speech-dispatcher-flite speech-dispatcher-espeak
The following NEW packages will be installed:
  espeak-ng-data gir1.2-atspi-2.0 libao-common libao4 libatk-adaptor libbrlapi0.7 libdotconf0 libespeak-ng1 liblouis-data liblouis20 libpcaudio0 libsonic0 libspeechd2 orca python3-brlapi python3-louis python3-pyatspi
  python3-speechd sound-icons speech-dispatcher speech-dispatcher-audio-plugins speech-dispatcher-espeak-ng xbrlapi
0 upgraded, 23 newly installed, 0 to remove and 0 not upgraded.
Need to get 11.1 MB of archives.
After this operation, 55.5 MB of additional disk space will be used.
Do you want to continue? [Y/n]

Hope this help.

sec4lib

2020-08-23 02:13

reporter   ~0013312

espeakup package issues ?
" This is pretty late, but at some time during 2018 some kind of fix was made to the ALSA driver, and speech using the default espeakup in conjunction with speakup no longer stutters.
This renders my solution, detailed in this article, no longer necessary. I leave this article here and the code downloads where they are, for reference, and in case problems re-appear."
http://www.raspberryvi.org/stories/stuttering.html ( Site last updated: 2019-03-23 23:06 by Mike. @ https://github.com/cromarty/

Issue History

Date Modified Username Field Change
2020-04-18 13:18 zersiax New Issue
2020-04-28 10:32 zersiax Note Added: 0012687
2020-08-23 01:44 sec4lib Note Added: 0013311
2020-08-23 02:13 sec4lib Note Added: 0013312
2020-12-01 10:48 g0tmi1k Priority high => normal
2020-12-01 10:50 g0tmi1k Severity block => minor