Hi,
Here is a summary of what happened:
In early September 2026, the original poster (of this thread) submitted DECtalk 01 add-on to the store for processing. I reviewed the code, and found that the add-on included a ROM dump of the DECtalk 01 model (circulating on the internet, by the way). I knew that people were posting ROM images of old speech synthesis hardware on some forums (or links to these images) and some have created SAPI5 ports of some old synthesizers. Then I learned of existence of a DECtalk add-on with the documentation stating that you (the user) needed to supply ROM image files and acknowledged DEC/Fonix as the copyright holder of ROM image data (if I read the docs correctly).
How did the add-on under discussion came to be? Wel, our friend cloned the NVDA community add-on template, forked it, then committed changes on top and renamed the repository. What changes? The already circulating DECtalk add-on without ROM files, then our friend added actual DECtalk 01 ROM image files. Shortly after, the DECtalk add-on was submitted for store registration approval.
When the DECtalk add-on submission was received, I reviewed the submission (code included) and pointed out the following notice from the add-on source code:
Ships no ROM bytes -- the DTC-01 firmware is DEC/Fonix copyright. The
user supplies their own dump; emu/rom_loader.py recognises the chips by
content hash (so filenames don't matter) and refuses to start otherwise.
See DESIGN.md §0.
Like I mentioned, Fonix ceased operations a few years ago. Technically, DECtalk and related technologies are abandonware unless someone with an actual copyright claim as of 2026 shows up and explains what is going on. Due to the unclear status of DECtalk and related tech, people have been working on activities such as dumping ROM images so they can be emulated, attempt to reverse engineer and/or decompile the speech synthesis materials, and do other things. According to some sources, the source code for DECtalk and related technologies was opened not long ago (if done by the original or current copyright holder for various reasons, then the terms of the source code distribution must be checked; if done by someone else, then we need to figure out how and why).
Copyright laws differ around the world. There are international treaties on copyright, and some countries (including United States) provide exceptions. While anyone can submit add-ons including speech engines, the fact the submission materials such as add-on data gets hosted on a source code repository service offered by an American company and the add-on is for a product where the place of origin is Australia complicates the scene. This came to light when I asked Google TTS submitters about copyright law differences for United States (GitHub/Microsoft, Google), Australia (NV Access), and the place of residence for the submitter (Vietnam) back in July 2026.
By the way, we do have several speech engine add-ons available from the NV Access add-on store. Some do include voice data (created by the authors themselves), and others do not include voice data but directs users to download the data either from within the add-on or from websites. For speech engines, the biggest question is, "do you have permission from the speech engine manufacturer to submit the add-on or have distribution rights?"
Cheers,
Joseph