Google might be ahead in some conversational things, but functionality wise, Alexa is way better.
Probably has something to do with being on the market first, but developing a skill is a much better experience than it is with Google (imo). I think that's a big reason why.
I've developed a number of skills for some big companies, and I'm fighting Amazon right now about getting my open source one published (https://github.com/m0ngr31/kodi-alexa)
I am concerned about the IAM policy you're choosing - Administrator access should basically never be given out or used. I would highly recommend figuring out what your actual dependencies are and restricting the policy to just what's necessary.
Yeah, I run it with the bare minimum, but I kept getting confused people not doing it right, so it was just easier to instruct full admin access.
If Amazon won't approve my skill (they have an issue with Kodi after the fire tv stick debacle), I have a hosted version that won't require setup like there is currently, so it's just a temporary solution.
I've seen some github repositories have a button you can click that supposedly instantiates an entire cloudformation stack. I haven't ever clicked on one, so I don't know how smooth the experience is, but it's arguably better than "go learn about terraform" or "click here, there, that, here, type a policy name, paste this policy in this box, click save".
According to the 'feedback' I received when I submitted the skill, they won't publish anything that enables or supports piracy. Seems they have a pretty negative view of Kodi.
I've responded and asked to have the issue escalated and posted on the dev forums, but the support is pretty sad, despite the otherwise good documentation and developer experience on the platform.
Could not disagree more. We have and echo since it came out late 2014 and now several Google homes. The Google home is just a lot more funtional. The obvious answering questions and has mapping built in but it is also foundational. The GH supports natural language for most things where the Echo requires rigid language or basically commands you have to memorize to use.
My wife clicks a photo on her iphone and without touching an additional button walks into our family room later and will ask for fine details in photos and the TV turns itself in, input sets and the photo in 4k appears.
We also have a 4k Chromecast and this is just not possible with the Echo.
But the cool part setting it up was just buy, plug in and log in and that is it. Wife already used Google photos on her iPhone.
We started with the Echo but now have switched to Google Homes and unless going to do a lot of shopping on Amazon can not see any reason to get an Echo over the Google home any longer.
Oh this is fantastic, I'm going to deploy it when I get back home. Thanks for opensourcing it!
One question: Which sections in the kodi.cfg are necessary? I already have Kodi configured and don't want to just replace the file, or spend time changing options your skill doesn't need.
I find 3rd party skills a painful experience to use.
Having to open a skill like an app before asking what I want Alexa to do feels completely wrong. It leads to me basically never using any third party skills.
If I wanted to open apps I’d use my phone. I want something more natural.
Probably has something to do with being on the market first, but developing a skill is a much better experience than it is with Google (imo). I think that's a big reason why.
I've developed a number of skills for some big companies, and I'm fighting Amazon right now about getting my open source one published (https://github.com/m0ngr31/kodi-alexa)