TTS file on S3 requires use of generic s3 endpoint?


#1

Hi all. I recently setup webcore on my smarthings hub. Many thanks for this amazing engine. I setup some TTS to a dedicated Sonos One speaker after my 7 year old daughter decided to sleepwalk right out of the house one morning at 1:08am. Webcore worked like a charm and allowed me to create some nice logic to turn on lights and make TTS announcements based on motion sensors and door sensors. Today however, the “The front door is open” announcement which fires when the front door opens died mid stream. It has not recovered.

I went into the piston and re-saved on my way back out. -> nothing changed
I went into the piston and changed the TTS announcement to “The front door is NOW open” -> nothing changed (no announcement)

The logs in the IDE show:
setTrack(https://s3.us-east-2.amazonaws.com/smartapp-media/polly-tts/salli/the_front_door_is_now_open.mp3, null, 0 B})

However, if I try to open that url it gives me an error stating that I have to use the specified endpoint…which is s3.amazonaws.com. When I replace the s3.us-east-2 with s3.amazonaws the file downloads and I can play it on my computer.

I am confident that you all know this already and that the redirected url shows up in the IDE logs…but, I am hoping that in face amazon changed something and there is a slight code tweak needed in webcore. :slight_smile:

Thanks for your help.

Jason


#2

This is quite a major situation, TTS “speak text” is broken it seems. i don’t know if it’s webcore or the ST hub, all i know is TTS speak text no longer works, as of 12 hours ago. i have pistons set up for Samsung speakers AND Google homes, they have both stopped working. Sending url radio streams still work though.

║Error while executing physical command Zzzzzz M5 Kitchen.playText([ok im just wondering if this is working now, it really should unless i’m totally confused, 47]): groovy.lang.MissingPropertyException: No such property: protocol for class: physicalgraph.device.HubMultiAction


#3

Instead of Speak Text have you tried the action “Speak” instead? That’s always worked better for me than “Speak Text”.


#4

I have not. As of this morning, my open door announcements are working. I am beginning to suspect (for no good reason other than guessing) that this all may be related to the pending firmware update today.