-
Notifications
You must be signed in to change notification settings - Fork 0
Description
Now many modules due to multiple interfaces look messy, esp. for understanding the API.
Should we split audio-speaker (and others) to
pull-audio-speakerfor pull-streamstream-audio-speakeroraudio-speaker-streamfor node streamaudio-speakerfor plain function- [
audio-speaker-clifor command-line?]
?
That would make API entrance easier, less single-package maintenance efforts, better docs, better isolation/independence, less weight of a single package, less naming conventions issues (like direct/stream/pull/etc), make for better tests (no need to cross-test stream versions), resolve naming issues (some audio-* modules are not streams already), allow for grouping stream-* and pull-* packages together to organize stacks, e.g. we could separate them in docs. Some people don't use streams, others don't use pull-streams, others don't know what streams are but want to output audio. I'd guess these are separate groups of users.