@nathan7, this is a known issue. You can currently only link one Google Home user to a TRIGGERcmd account, so if you use the Voice Match feature so multiple Google accounts use the same Google Home, you run into this issue.
It's on my list of things I need to solve. Ideally when you link one Google account via the Home app, and there's already an account linked, TRIGGERcmd doesn't replace the original link, it adds another link.
@zipster1967, ok, please try deleting your C:\Users\trent.TRIGGERcmdData folder. I've seen Avast cause problems in the past, and I suspect it put your data folder in a broken state. If you delete the folder, TRIGGERcmd will re-create it. I think once the data folder gets created, you can turn Avast back on.
@zipster1967, I think you do need to re-install TRIGGERcmd agent, but before you reinstall, will you please try disabling your Antivirus software? Some Antivirus software breaks the installer system TRIGGERcmd uses.
Also, can you tell me what Antivirus software you use? I'd like to see if there's a way to make the installer compatible.
Thank you @david-jiménez-lamas! I'll try to get this done soon. Expect a couple weeks though. Do you have Alexa or Google Home? They have different processes for getting approved.
@jorgee, have you tried deleting your c:\users\(your user)\.TRIGGERcmdData folder? If you do that, then close and re-open TRIGGERcmd, it might fix it.
Another possibility is your Antivirus software is blocking it. You could try turning that off temporarily while TRIGGERcmd runs for the first time.
@fernando-mateos, no problem. Same here.
@cmac, I tried waking up my PC from my mac, and I came up with these tips for sending the wakeonlan "magic packet" from a TRIGGERcmd command on a mac:
First, make sure you've enabled Wake on LAN in the PC's BIOS, like so:
Get your PC's network interface's mac address. Google how to do that if you don't know. Also, it has to be a wired connection, not wireless. You probably already knew that, but just in case.
On the mac, do the brew install wakeonlan command to get it installed, then open a terminal and try it using a command like this:
wakeonlan c0:7c:d1:a1:89:f8
… where c0:7c:d1:a1:89:f8 is the PC nic mac address.
If that works, you're golden. The next thing is, in the TRIGGERcmd command on the mac, you have to specify the full path to the wakeonlan binary. Brew installs in as /usr/local/bin/wakeonlan.
So you command field should look like this:
/usr/local/bin/wakeonlan c0:7c:d1:a1:89:f8
YMMV. In my case, I couldn't turn my computer on if it was off. I could only wake my PC up from sleep mode.
@atmosphere, I'm sorry but right now the TRIGGERcmd installer only supports one TRIGGERcmd account per Raspberry Pi.
But, theoretically you could maybe run two instances of the TRIGGERcmd agent on the Pi, as two separate Linux users, and give those two users sudo access if they need to run things as root.
If you run triggercmdagent --console it will run in foreground mode like so, and store two separate commands.json files in the .TRIGGERcmdData subfolder of their respective home folders:
pi@raspberrypi:~ $ triggercmdagent --console
Run installdaemon.sh to install the triggercmdagent daemon so it runs during boot
Daemon install: false
No token exists. Login to request one.
prompt: token:
The trick would be in getting the two agents to run after a reboot.
@karl-g, I see you have a space in the path to your batch file, so you'll need to add quotes around it like this:
"C:\x10 shorts\bedroom heater on.bat"@karl-g, please uninstall and re-download from the site. I took the new version I created last weekend down, and put v1.0.19 back because you and one other person reported this issue.
@karl-g, please uninstall and download again. I removed version 1.0.20 and put 1.0.19 back until I can figure out why people are getting that error.
@jdeuel, just in case I broke something in this latest 1.0.20 version, I took it down and put 1.0.19 back, so please try downloading that version from the normal link on the www.triggercmd.com website.
I'm not having any trouble with 1.0.20 on my computers, but two people reported an issue, so I took it down.
If it still doesn't work, another thing you could try is temporarily disabling your Antivirus software. I've seen certain AV software cause issues in the past.
@kaiontour, unfortunately I only have so much control over what Alexa will recognize, which is why I gave the user the flexibility in that voice field to pick something that Alexa will recognize. You could try "restart kodi" instead of "kodi restart" if that works better.
@jdeuel, please try deleting both your commands.json and commands.json.backup files, the re-run the TRIGGERcmd agent. It will re-create those files when you re-run it using the default commands.
@snantz, I haven't added that voice response feature yet, but when I do, this is how it will work:
If you put {{result}} in the Voice Reply field of your command (like this), Alexa or Google Assistant will wait up to 5 seconds for a result to be sent from the computer, and include it in the voice response.
If you upgrade to the latest version of the agent (as of 1/6/19), you'll now find a new sendresult.bat or sendresult.sh file depending on your computer's operating system. You can send the result of your triggered script by adding this to the end of it:
Windows:
%USERPROFILE%\.TRIGGERcmdData\sendresult.bat "My command result"
Linux or Mac:
sh ~\.TRIGGERcmdData\sendresult.sh "My command result"
Today the sendresult script just logs your result in the command's Runs list, but this is the ground-work for this voice feed back feature you asked for.
@gutopa, I've been thinking about your question and how I'd implement a solution. Please look here for what I'm thinking.
@snantz, commands.json gets overwritten with a backup of commands.json if the TRIGGERcmd agent finds invalid json in the commands.json file. One common mistake is to leave a comma on the last line.
@snantz, this is the second request for something like this. I've been thinking about how I would implement that.
I'm thinking I could give an option you could enable on each command to tell Alexa or Google Assistant to wait a few seconds for a response from the computer. If the computer sends a response, then Alexa or Google Assistant would speak the text in the response. I'm thinking the response could be sent with a curl or tcmd command.
What do you think?