i don’t think there is any mechanism built in to support this… but, you could use MMM-RemoteControl on the server side that accepts api requests.
then on the RPI you would have to write some background tasks to interact with the local devices
and send requests.
you ‘might’ be able to inject those handlers into the MM front end by modifying the index.html
@reb those characters are called the BOM (Byte Order Mark)… added by advanced word processing editors…
u can use Notepad++ to remove them too…
u need to change editors to make sure you don’t add them back accidentally…
u have to write some additional communications process.
in one of my implementations, i write a file in the outside script, and my MM modules looks for the file and then processes it… which then sends notifications to other modules…
you could also use tcp/ip sockets, or websockets, or … there is no single builtin way
@klaus add debug: true, to the mmm-voice config.js settings
then there should be a debug.log in the mmm-voice folder
what does it do? I just ported over my pure node/angular/javascript code to mm…
took some thinking about how to separate the existing code to fit the MM execution model…
in general on MM you supply fragments of a page to be shown in some location on the MM page… you can also do a whole page in an iframe…
@bjoern new installation of what? MM or PI OS?
@tuicemen the motion project can also capture video and stills. I don’t think it can stream live.
@sdetweil also, i don’t think u can send keystrokes
@madscientist others have mentioned commenting out the updater module