Speakup Text Mode or Orca Graphical Mode for Rocky Linux Installs on Uncertain Hardware

Question

Rocky Linux offers two documented accessibility paths in its Anaconda installer, and a deployment plan for blind and low-vision users has to commit to one before the capabilities of every target machine are known. The goal is to standardize a single installation procedure that stays dependable across mixed hardware.

The text-mode console can be made accessible with the Speakup screen reader, but Speakup is not active by default; it must be requested through a kernel parameter such as speakup.synth=soft at the boot menu, or built into a custom kickstart file. The graphical GTK installer exposes Orca and high-contrast themes, yet Orca only becomes available once the graphical session starts, so the path depends on working video drivers and more memory. Documentation also frames Speakup partly as a legacy option, and synthesizer detection can fail silently on some USB keyboards and virtualized consoles.

That leaves the choice unresolved when RAM, video support, and console type vary:

Which installer mode is the safer default for screen-reader users when hardware capabilities are uncertain?
Is there a supported way to make Speakup available during installation without editing boot entries on every run?
At what practical minimum of memory and console support does the graphical Orca path stop being viable?

Tasadduq Burney · Answer

Answer
The safer default for screen‑reader users when hardware capabilities are unknown is the Speakup text‑mode installer.
Why Speakup is the safer default

Speakup runs from the early console, needing only a functional VGA text mode or framebuffer.
  It works with as little as ~256 MB RAM and does not depend on video drivers or a graphical session.
  Orca requires the GTK/GNOME installer to start, which needs a working KMS framebuffer and typically > 768 MB RAM; without those resources the graphical path may fall back to text or fail to start Orca.

Making Speakup available without editing boot entries each time
Add the kernel parameter to the default boot loader configuration so it is applied on every boot:

Edit the GRUB defaults file:
  sudo vi /etc/default/grub
  Append inst.speakup (or a specific synth, e.g., speakup.synth=soft) to GRUB_CMDLINE_LINUX:
  GRUB_CMDLINE_LINUX="... inst.speakup"
  Regenerate the GRUB configuration:
  sudo grub2-mkconfig -o /boot/grub2/grub.cfg

Alternatively, embed the same parameter in a custom boot ISO or Kickstart file that is used for all installations.
When the graphical Orca path stops being viable
Orca becomes unreliable when either of the following is true:

Available RAM is roughly less than 768 MB** (the installer’s GUI may not start or may be extremely slow).
  No working video driver/KMS framebuffer is present (e.g., headless servers, certain USB graphics adapters, or virtual consoles without /dev/fb0). In this case Anaconda falls back to text mode and Orca cannot launch.

If you can confirm that target machines have at least 768 MB RAM and a detectable framebuffer (/dev/fb0 or fbset shows a device), you may consider the Orca path; otherwise stay with Speakup.
Verification steps (optional)

To confirm Speakup is active: boot with inst.speakup, then listen for speech and run dmesg | grep -i speakup to see the driver load.
  To check framebuffer presence: ls -l /dev/fb0 or fbset.
  To test RAM: free -m during the installer shell (Ctrl+Alt+F2).

Missing diagnostic detail that could change the recommendation: please confirm the minimum RAM and presence of a framebuffer device on the hardware you intend to deploy.

Speakup Text Mode or Orca Graphical Mode for Rocky Linux Installs on Uncertain Hardware

1 answer

Answer

Why Speakup is the safer default

Making Speakup available without editing boot entries each time

When the graphical Orca path stops being viable

Verification steps (optional)

Your perspective belongs here.

1 question comment