MMM-AssistantMk2

[A Former User]

@themc
If you have a module to play mp3 and it is able to be activated by notification, it could. But adding some specific job to Assistant itself, I think it is not so good idea, because each circumstance is different.
So have you a module which can play mp3? I can make gaction sample for it.

[dvbit]

@sean thanks for all the info.
Just saw this after sending you request for info.
will search better / more in the future :-)

[dvbit]

Looking for a solution for my vertical setup and browsing the file system I have seen that the cardsare stored in a temp_screen html file in the plugin folder.

changing there font size from 40px to 20px makes already a big diff (no scroll bars… in the cards)

is htere a way I can intercept that file and edit before presentation?

that would already do a good trick :-)

[A Former User]

@dvbit
I’ve already done that. screenZoom is that.
But anyway, That screenOutput was made for SmartTV, so it was optimized for Landscape mode. And unfortunately, the screenOutput is uncontrollable static HTML itself. There is no enough room for control.
I wish the next version of Google SDK could provide structured data instead static HTML for screenOutput.

[dvbit]

Ok. Tx will use that parameter then :-)

[mariecbilodeau]

@sean here is my complete log
https://www.dropbox.com/s/7l4wd8xtb3t9odk/error log?dl=0
thank you for your help :)

[mariecbilodeau]

I got it working with your solution in the troubleshooting section!!! thx

[ToastedToasty]

@Sean Hi Sean I’ve been trying to setup MMM-Assistant however after renaming my client-secret.json file to credentials.json and running the auth_and_test.json the OAuth2 page to get my token gives me an error 500, any ideas?

[A Former User]

@toastedtoasty Usually, 500 error means your google project has something wrong. I suggest you remove your google project and recreate again. (I doubt you might have two or more projects and have some confusing about selecting them)

[dvbit]

@Sean
I have all setup correctly and no error with the configs.
I think I have understood the correct working of transcriptionhook but it does not fire the related actions.
Can you confirm it is like this the logic:
I setup a keyword in the config (unicorn) in the example you provide.
Notification trigger catches the word if it is spoken after the hotword.
The related action is fired.

Unfortunately nothing is triggered . Also : should I see some feedback in the console windows if the hook triggers ?
Thanks as usual for the support