[katsiris]

Woohoo! I finally got Alexa working by going back down to Python 2.7 and re-deploying! Thanks all for your hard work on this, it is really excellent in action! Has anyone seen the announcement for Alexa on Plex? They have a pretty nice implementation with suggestion logic (by genre? start a new show? Watch the next episode of x?) that would be fun to implement. Would that all be done in utterances?

[bill_orange]

Woohoo! I finally got Alexa working by going back down to Python 2.7 and re-deploying! Thanks all for your hard work on this, it is really excellent in action! Has anyone seen the announcement for Alexa on Plex? They have a pretty nice implementation with suggestion logic (by genre? start a new show? Watch the next episode of x?) that would be fun to implement. Would that all be done in utterances?

After seeing the announcement for Alex on Plex, I tried it. I was not impressed. Here are the pros and cons:

Pros;

Excellent speech recognition.

Good set of utterances.

Large user base and wide acceptance.

Good documentation.

Cons:

Appallingly slow. I loaded the server on a PI3 and it ran like molasses. In two hours it had not scanned in all the "A"s in my library. Kodi scans the whole library in 45 min.

They want money. You need a paid subscription for the client on the Raspberry Pi 3 to be fully functional. There are lots of restrictions unless you send money.

Speech integration only works on certain clients. It will not work on my Samsung TV.

In summary, It took me the better part of a morning to get all the permissions issues resolved on the server so the library could start scanning. Then I discovered that without sending money the PI client would not work with speech integration, The TV client would not support Alexa at all. By dinner, I could stream most of the "A"s to my phone!

I uninstalled it and gave up. I did learn a bit about Linux permissions, however.

[Farquea]

May be a stupid question but once I have this all up and running (I do) am I ok to uninstall Python and delete the Kodi-Alexa-Master file from my computers, there's no link now needed between these or will be at any point, thinknig updates etc?

[jingai]

May be a stupid question but once I have this all up and running (I do) am I ok to uninstall Python and delete the Kodi-Alexa-Master file from my computers, there's no link now needed between these or will be at any point, thinknig updates etc?

For updates you will need to keep these things around.

[bill_orange]

Is the Eject function not fully implemented? I doesn't seem to work.

If I say " Alexa, ask kodi media to eject", I get the same beep as I would get if I said " Alexa, ask kodi media to feed the cat".

In wsgi.py, I changed the answer = " " for this function to answer = "I don't see one. Maybe one is hiding under my stubby little legs" to see if the line was uttered. The line is never uttered.

I also tried commenting out kodi.SystemEjectMedia() in case it was crashing inside the function call. No effect.

So, I'm confused. I just don't know enough yet to figure this out.

[jingai]

@bill_orange, sometimes Amazon will intercept certain things at their whim...

I just tried, and "Alexa, ask Kodi to eject CD" worked. DVD and media did not, but there's not a whole lot we can do about it if Amazon won't even pass it to the skill.

[bill_orange]

@bill_orange, sometimes Amazon will intercept certain things at their whim...

I just tried, and "Alexa, ask Kodi to eject CD" worked. DVD and media did not, but there's not a whole lot we can do about it if Amazon won't even pass it to the skill.

I will try it with the eject CD. How odd. How could Amazon pass one and not the other?

[jingai]

Because they reserve certain keywords/phrases, apparently, but they don't disclose that to us.. nor do they tell us what magic is required to bypass it, if any.

[bill_orange]

No Luck with "eject CD" here. After I upload to Lamda does it take time for the new version to propagate? If so, I will try it again in the morning.

[digiltd]

No Luck with "eject CD" here. After I upload to Lamda does it take time for the new version to propagate? If so, I will try it again in the morning.

No, it should be available a few seconds after it has been uploaded.

Try replacing the sample utterances for EjectMedia with these:

Code:

EjectMedia eject
EjectMedia eject audio disc
EjectMedia eject blu-ray
EjectMedia eject blu-ray disc
EjectMedia eject c. d.
EjectMedia eject compact disc
EjectMedia eject d. v. d.
EjectMedia eject disc
EjectMedia eject media
EjectMedia eject the audio disc
EjectMedia eject the blu-ray
EjectMedia eject the blu-ray disc
EjectMedia eject the c. d.
EjectMedia eject the compact disc
EjectMedia eject the d. v. d.
EjectMedia eject the disc
EjectMedia eject the media

According to the docs you need to include full-stops (periods) after letter abbreviations.

Though even with that change I couldn't get it to work, Alexa is always hearing "check the cd", "check media" instead of "eject". Quickly saying "check" and "eject" repeatedly to yourself reveals how similar those two words actually sound. But as I don't actually have a disc drive I will let you have a test and feedback what you discover

You could always try adding your own utterances e.g. "remove the c. d." or "open the disc drawer". Though "eject" is the most obvious choice and we should try to get it working.

[jingai]

@digiltd, though that doesn't explain why "eject media" doesn't work. Also, have you tried full-stops with capital letters and no spaces?

Though I'd have a hard time believing Alexa can't work out that a word in all caps is an acronym..

edit: Looked it up instead of being lazy like I'd planned.. lol.. the docs say the way we had it is right:

For better recognition, acronyms and other phrases involving spoken letters should either be all caps (“HD”) or separated by periods (“h.d.”). Using lowercase for initialisms may lead to unreliable recognition since the spoken form may not correctly be detected.

[bill_orange]

No Luck with "eject CD" here. After I upload to Lamda does it take time for the new version to propagate? If so, I will try it again in the morning.

No, it should be available a few seconds after it has been uploaded.

Try replacing the sample utterances for EjectMedia with these:

Code:

EjectMedia eject
EjectMedia eject audio disc
EjectMedia eject blu-ray
EjectMedia eject blu-ray disc
EjectMedia eject c. d.
EjectMedia eject compact disc
EjectMedia eject d. v. d.
EjectMedia eject disc
EjectMedia eject media
EjectMedia eject the audio disc
EjectMedia eject the blu-ray
EjectMedia eject the blu-ray disc
EjectMedia eject the c. d.
EjectMedia eject the compact disc
EjectMedia eject the d. v. d.
EjectMedia eject the disc
EjectMedia eject the media

According to the docs you need to include full-stops (periods) after letter abbreviations.

Though even with that change I couldn't get it to work, Alexa is always hearing "check the cd", "check media" instead of "eject". Quickly saying "check" and "eject" repeatedly to yourself reveals how similar those two words actually sound. But as I don't actually have a disc drive I will let you have a test and feedback what you discover

You could always try adding your own utterances e.g. "remove the c. d." or "open the disc drawer". Though "eject" is the most obvious choice and we should try to get it working.

I am at an impasse in editing the skill. Any attempt to save results in:

Error: Antonin Dvorak: Symphony No.9, "From The New World"/Carnival Overture contains invalid characters. Values for a custom slot type cannot contain double quotation marks.

If I correct the error and try to save again it picks another song or album and will not save or let me save an edit.

I don't want to delete the skill and start again. Suggestions?

[jingai]

Remove the quotes in your slot items. The slot generator will eventually do this for you, but for now, just do it by hand.

[bill_orange]

Remove the quotes in your slot items. The slot generator will eventually do this for you, but for now, just do it by hand.

I wish I could do it by hand, but there are "" in both MUSICSONGS and MUSICALBUMS and it won't let me fix and save the first because there are still "" in the second. Tha's what's preventing me from saving. It looks like a catch 22 unless I am missing something.

[jingai]

Yes, you're missing that there are two different save buttons. The one at the bottom of the page is to save the skill itself, while the one underneath the slot item list is to save the Slot. Edit each list one at a time and use the latter to save the slots, then save the skill with the bottom button.