Simple Speech to Hypermedia System

1. speech-to-text and a REST microservice over todo.txt, or to vimwiki functions in bash. Then you can do lazy encryption.
2. Use the HTML5 Media driver for audio and capture record handles on textblocks. Inherit linking schemas from existing doctypes. You could use RSS too. Point being todo.txt and .wiki
3. files are just text. Infinitely portable. You can just save to git repos and use github/gitlab/bitbucket as your database, if it is strictly personal. Set up folder watchers in bash that check last modified date on project folder with diff of cached folder date, then reset REST service in git post-push.
4. If you want access rights map RESTful commands to chown statements in bash that update a custom group versus public. They accessthe file if they pass in a cgiparam that you specify is the password. Use bitcore's HD wallet to use private key generation for groups then you could share with 2M people per peer group. Scales.
5. Now broadcasting on all subnets in the languages.