Gemini text-to-speech 'podcasts'? - I had the thought recently that it would be nice if I could queue up a playlist of people's gemlogs and listen to them through a text-to-speech service. Since gemini is so text-centric, it seems like this should be able to work very well. It would especially be nice if this kind of support could be built into a mobile gemini browser with basic play, pause, and skip controls.
What are your thoughts? Does something like this already exist? Would you use a feature like this for listening to people's blogs and essays?
Aug 04 · 2 months ago
2 Comments ↓
🦎 bluesman · Aug 04 at 15:03:
That's an interesting idea and it prompted an experiment.
The next version of Alhena can send pages to a printer. This means you can create PDFs if you have an appropriate driver. I output your post to PDF and used the PDF viewer installed on my ancient version of Windows to read it back. It worked but it was a bit slow. This may be configurable or something a newer or different viewer could do better.
👾 jecxjo · Aug 04 at 23:14:
A simple hackable way would be to use something like espeak to pipe the text out into an mp3. You'd need a curl-like client or netcat.
#!/bin/sh
SERVER="$1"
URI_PATH="$2"
printf "gemini://${SERVER}${URI_PATH}" | \
openssl s_client -quiet -ign_eof -connect "${SERVER}:1965" 2>&1 | \
espeak -s 140 -v en-us --stdout | \
ffmpeg -y -i - output.mp3
Source