question

+1 Things like custom slot types and the built-in intents are additional examples where a visible roadmap might've been helpful in preventing people from implementing things that are now obsolete.

Matt, I agree about the need for a roadmap. I'm not not quite sure what the rationale is for not having one. I had considered some design decisions in some recently skills that, had I kept them in, would have been killed by the new user id handling. (I just happened not to do it that way, but it would have caused some trouble.) So +1 for a road map. And, even if it's not publicly announced, just a quick note to a dev about upcoming native skills if the cert team sees that the dev's work is going to be in vain -- that seems like a courtesy.

I would like to see a roadmap too. But I also understand two things that might explain the lack of one: 1) Amazon are paranoid about competitors. If they see this as a highly competitive field, then they might be "giving the game away" by saying what they are doing next. Their competitors could then specifically develop those features to be sure they had feature parity. 2) 3rd party developers are not where Amazon sees high value for Alexa. Well all well know the advantage the "built in" features have over 3rd party skills. If you add up the number of Amazon names posting to this forum and divide by how much time they spend posting, it's clear they have other duties. Same for those doing certification. So, what are they doing? Something Amazon thinks is more important than supporting 3rd party skills. So, ultimately, [b]we [/b]definitely need a roadmap. We can't make a business out of this without one. We can't avoid wasting our time without one. We can't be anything more than hobbyists without one. But our goals are not necessarily aligned with Amazon's goals. I hope that this is a "point in time" statement, though. In the same way that 3rd party content really pushed Facebook through the roof, I think that 3rd party content is a huge value-added to Alexa. This may become more apparent over time, and we might see Amazon shift more towards treating us as business partners, rather than consumers. Time will tell.

Hey Matt, I really liked your EchoTTT skill. Even with IFTTT's trigger function, I'm not able to set lights to a specific color any longer without making a recipe for each color. In addition, I had created other automations using as RasPi that no longer function. I have a AWS Lambda acct. and I was wondering if I could use your lambda code for my own personal function that's based on EchoTTT. I assume that if I customize my own endpoint, I can still use the functions? Thanks in advance, and your EchoTTT project is by far the best I've come across! J

>your EchoTTT project is by far the best I've come across! Thanks! :) Unfortunately, Amazon has made a habit of rejecting all the really useful skills I've created, while approving only a stupid Guess The Number skill. Bummer. Unfortunately my code was tied to a Dynamo database, and running off a VM in the Google Cloud. I'm not sure how useful my code would be, considering you would need to duplicate my environment. But I will put it up on github soon so you can do with it whatever you wish, but you'll at least need to setup your own server or create a Lambda function.

Sounds good Matt! I have an AWS acct and I've been playing with lambda a tad as well. My coding skills are limited, although I find python a bit easier to work with, so I managed to tinker some code to input multiple values into a pycurl command and get that to post to IFTTT. Since I'm using pyCurl to post, I'm having issues getting it to work via AWS lambda. Maybe once you get your code up on github, I'll be able to glean a few insights and go from there. Much appreciated and keep up the good work! J

Okay, here it is on github: https://github.com/matt-kruse/echottt See what you can do with it :)

Very elegant code there Matt! I altered AtoIFTTT's code to function the way I want on AWS, yet I was missing a way to convert numbers to digits. The words2numbers.js code is particularly helpful there. Thanks again! I'm sure the weekend code warriors, aka amateurs like myself, appreciate the help! J

+1 on need for roadmap. This is one of the major reasons I haven't invested in further skill development after getting my first four certified. I don't want to put in the work and time when it's possible---likely, even---that Amazon's own team is already working on a native version of whatever I've got in mind. I've looked at some APIs but ultimately stepped back because I know Amazon's own team may be working directly with the company behind the API to release a more seamless version of whatever I'm considering as a possible skill. ='/

Great point about built-in functions (e.g. I thought they would have a stock quote built-in and was surprised when Fidelity released a skill for this) +1 for roadmap