I've had this conversation with a couple of OSS developers and the answers always leave me very uncomfortable.
The problem is how does one live by OSS principals when essential tools are vehemently closed and the barriers to replacements are decade scale and no one is working on them? The problem I refer to is the use of speech recognition as a tool for dealing with upper extremity disabilities. There is only one vendor for continuous speech large vocabulary recognition. There is maybe one or two universities in the world conducting research into speech recognition. All the open-source toolkits are hampered by design criteria (fixed grammar small vocabulary) and there is no corpus sufficient to build acoustical models. Recognition engines are multimillion dollar efforts to build and corpus collection is even more expensive. Speech recognition also requires very specialized knowledge and the people skilled in the art are owned by industry. Therefore, if a rational person would assume that OSS speech recognition is not coming anytime in the near future, maybe not even in my lifetime. A rational person would also assume that part of the way to tackle the problem is to nibble at the edges from the application side to the recognizer side, gradually increasing the availability of OSS components so that the disabled person can minimize their dependence on proprietary or closed source applications. A lot of us disabled programmers have done a good job the nibbling around the edges but there's a lot of cases where we don't have the knowledge and need the help of project related people for example, Emacs integration mode with NaturallySpeaking (VR-mode) doesn't work right. It is incredibly fragile and breaks apparently at random. When I asked for help from various Emacs wizards to help keep it up-to-date and maybe even integrated into Emacs source in the hopes that it would be less likely to break, I was told there was no chance of help because it was linked to a proprietary package. That doesn't leave us in a very good place because if that attitude persists from the ideologically pure, disable users have a shrinking number of open-source applications they can use because, the users require the use of a proprietary package. How does one deal with the real world issue that disabled users will need proprietary packages integrated with open source applications to keep them from being forced into using 100% proprietary applications with no options? -- Ubuntu-accessibility mailing list Ubuntu-accessibility@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-accessibility
