EvoControl - custom Alexa skill for Evohome

[Stevedh]

I'm pretty sure that there's a hook script you can have certbot run when you do the auto-renewal and you can do your magic in that script. The certbot docs will reveal all.

Yes, if you did a full unlink/re-link, the default starting state of the enableHGI parameter is true unless manually over-ridden or automatically disabled due to comms fault.

EDIT: Just remember, you'd lose your scenes/groups and saved [BC] schedules if you did that

Thanks will look into the hooks when I get a chance.


HGI-80 via Domoticz is really cool, but pretty much only open to enthusiasts.

Just wondering how easy it would be to create a plug and play device probably linked to a cloud server somewhere to get around all the certificate stuff.
although I spose the market may not be big enough for that.... just musing.

[philchillbill]

Thanks will look into the hooks when I get a chance.


HGI-80 via Domoticz is really cool, but pretty much only open to enthusiasts.

Just wondering how easy it would be to create a plug and play device probably linked to a cloud server somewhere to get around all the certificate stuff.
although I spose the market may not be big enough for that.... just musing.

The HGI-80 stuff is definitely for nerds, but then again this forum is all nerds to begin with It's an immense pity that Honeywell does not expose the info collected by RF via the API. But it is what it is.

I already wrote a script that makes using ngrok set-and-forget for use with EvoControl/HGI. See https://smartskills.tech/ec/help/easyngrok.html

[jweaver]

This is almost certainly a stupid question, but I don't generally use Alexa much other than listening to the odd radio station, and am not used to "Stringing" commands together or controlling 'skills'.

On the Show, i understand how it works.... But this brings me to a quick question.. Is there a way to bring up "Evocontrol" without talking? Can you create a shortcut on the UI? There are times I want to look at something but don't wan't to keep saying "Alexa, open Smart Heat" every time.. (and I have to do this lots as I keep saying 'cancel' and closing the skill).

Anyway.. My question.. When Evohome is running, it seems to take voice commands just fine... And I thought I had it working on a Dot too, where i could say "Alexa, ask Smart Heat to set Study to 23 degrees".

But this morning, I can't get any of my dots to recognise this... I tried speaking faster.. Slower.. Clearer, but I can't make it work.. The only thing which works is "Alexa, ask smart Heat".. Wait for a response and then "set study to 23 degrees".

Invoking the skill first (alexa, ask smart heat) and then giving a command works fine... But I just want to understand why I can't string it into one single sentence and was 100% sure I managed it last night..

As a side note, at one point I realised I had added the word "temperature" instinctively (so "set study TEMPERATURE to 23 degrees") and that stops it from working, but I realised my mistake and dropped it. (is this something you might want to add, just to enhance the range of phrases it understands?)...

What am I doing wrong?

[jweaver]

One more question.. Is tehre any difference in using the "standby" activity? vs setting the whole house to 5 degress?

I used 'standby' last night and it was great.. But afterwards figured that I could achieve the same by just say "set entire house to 5 degress until 6am"... What does standby offer that "entire house" doesn't?

[philchillbill]

This is almost certainly a stupid question, but I don't generally use Alexa much other than listening to the odd radio station, and am not used to "Stringing" commands together or controlling 'skills'.

On the Show, i understand how it works.... But this brings me to a quick question.. Is there a way to bring up "Evocontrol" without talking? Can you create a shortcut on the UI? There are times I want to look at something but don't wan't to keep saying "Alexa, open Smart Heat" every time.. (and I have to do this lots as I keep saying 'cancel' and closing the skill).

Anyway.. My question.. When Evohome is running, it seems to take voice commands just fine... And I thought I had it working on a Dot too, where i could say "Alexa, ask Smart Heat to set Study to 23 degrees".

But this morning, I can't get any of my dots to recognise this... I tried speaking faster.. Slower.. Clearer, but I can't make it work.. The only thing which works is "Alexa, ask smart Heat".. Wait for a response and then "set study to 23 degrees".

Invoking the skill first (alexa, ask smart heat) and then giving a command works fine... But I just want to understand why I can't string it into one single sentence and was 100% sure I managed it last night..

As a side note, at one point I realised I had added the word "temperature" instinctively (so "set study TEMPERATURE to 23 degrees") and that stops it from working, but I realised my mistake and dropped it. (is this something you might want to add, just to enhance the range of phrases it understands?)...

What am I doing wrong?

Amazon is pretty adamant that skills can only be launched by voice command so unfortunately there's no way around that. There are 'widgets' that can be displayed only on the 15" Echo that might be able to do what you want, but Amazon's documentation on that is so poor that I don't even know where to begin. It's even unclear if they can be tapped to launch the underlying skill. But believe me, I'll be all over it when it becomes possible because I want this feature badly for many of my skills.

I can keep a session open when a skill supports APL and a device has a screen. However, on screenless devices there's 8 seconds to answer questions posed by Alexa or a skill is forced to exit. You can't just have an Echo Dot sitting for minutes with a blue ring waiting for a command for a specific skill. I can see the reasoning in that, because Alexa needs to be available for any command and keeping a skill open limits her to just the skill's commands. I could add an EvoControl command for screenless devices which would keep a session open for successive bouts of 8 secs. So you'd start on an Echo Dot with "Alexa, open Smart Heat". Your first in-skill command would be "Stay open". She'd say, "OK, what's next?". And you'd keep issuing commands within 8-sec windows back-to-back until you say "Exit". My MediaServer skill for Logitech Media Server works that way. The thing is, it's very unforgiving of any humming and hawing so I decided not to do it with EvoControl. That's the beauty of a screen. The skill Lambda in the cloud actually exits after each command is processed, but the screen markup stays put. Your next command relaunches the skill in the cloud, etc. But because the screen data stays displayed, you have the illusion of continuity.

The failure to understand one-shots starting with "Alexa, ask smart Heat..." is a random Alexa problem and not related to this skill. It happens on all skills. I like to joke at home that Alexa is PMS'ing today (how un-woke is that?). It will likely not happen later today or tomorrow. It never seems to affect Echo Show devices, only Dots.

I'll have to see if I can add the word' temperature' to the voice model because the model for that command is already pretty stuffed with a myriad of syntactic variations and there are limits.

[philchillbill]

One more question.. Is tehre any difference in using the "standby" activity? vs setting the whole house to 5 degress?

I used 'standby' last night and it was great.. But afterwards figured that I could achieve the same by just say "set entire house to 5 degress until 6am"... What does standby offer that "entire house" doesn't?

You can schedule an activity to start later whereas a "Set entire house..." command always starts now.

[jweaver]

Amazon is pretty adamant that skills can only be launched by voice command so unfortunately there's no way around that. There are 'widgets' that can be displayed only on the 15" Echo that might be able to do what you want, but Amazon's documentation on that is so poor that I don't even know where to begin. It's even unclear if they can be tapped to launch the underlying skill. But believe me, I'll be all over it when it becomes possible because I want this feature badly for many of my skills.

I can keep a session open when a skill supports APL and a device has a screen. However, on screenless devices there's 8 seconds to answer questions posed by Alexa or a skill is forced to exit. You can't just have an Echo Dot sitting for minutes with a blue ring waiting for a command for a specific skill. I can see the reasoning in that, because Alexa needs to be available for any command and keeping a skill open limits her to just the skill's commands. I could add an EvoControl command for screenless devices which would keep a session open for successive bouts of 8 secs. So you'd start on an Echo Dot with "Alexa, open Smart Heat". Your first in-skill command would be "Stay open". She'd say, "OK, what's next?". And you'd keep issuing commands within 8-sec windows back-to-back until you say "Exit". My MediaServer skill for Logitech Media Server works that way. The thing is, it's very unforgiving of any humming and hawing so I decided not to do it with EvoControl. That's the beauty of a screen. The skill Lambda in the cloud actually exits after each command is processed, but the screen markup stays put. Your next command relaunches the skill in the cloud, etc. But because the screen data stays displayed, you have the illusion of continuity.

The failure to understand one-shots starting with "Alexa, ask smart Heat..." is a random Alexa problem and not related to this skill. It happens on all skills. I like to joke at home that Alexa is PMS'ing today (how un-woke is that?). It will likely not happen later today or tomorrow. It never seems to affect Echo Show devices, only Dots.

I'll have to see if I can add the word' temperature' to the voice model because the model for that command is already pretty stuffed with a myriad of syntactic variations and there are limits.

Thanks for that.. its weird.. I literally can't get any of my dots to understand a single sentence today (ask smart heat to set study to 23 degree).. it either responds with "which zone".. or "I please say an Evocontrol command".... I must have tried 50 times and it doesnt' work.. Yet if I say "ask Smart heat" and then wait for the skill to invoke before giving the command, it works every time.. I think I will just stick with the latter method which works.

As for adding "temperature", you really don't need to bother.. It was just my misunderstanding.. I tried to talk to it how i ASSUME it worked and instinctivly threw in "temperature".. But after looking at the examples, I realised it wasn't needed.. I guess its nice to have incase you accidentally slip the word in.. But its hardly a big deal.

Its a real shame you can't launch skills by touch on the Show... The whole product just feels a bit useless really.. For example if you want a browser, you have to talk to launch it, even thought you are going to control it with your fingers.. But I guess they dont' want it to turn in a tablet??

[jweaver]

You can schedule an activity to start later whereas a "Set entire house..." command always starts now.

Ahhh.. I see.. So if I am leaving to go to work, setting the whole house, or invoking an activity will do the same thing? Thanks for that...

[philchillbill]

Ahhh.. I see.. So if I am leaving to go to work, setting the whole house, or invoking an activity will do the same thing? Thanks for that...

Whichever variant you use, remember to mention an until or duration. Otherwise, it's only until the next scheduled switchpoint reverts things. So if you're coming home from work at 5PM, add "... until 5PM" to your utterance.

[philchillbill]

Thanks for that.. its weird.. I literally can't get any of my dots to understand a single sentence today (ask smart heat to set study to 23 degree).. it either responds with "which zone".. or "I please say an Evocontrol command".... I must have tried 50 times and it doesnt' work.. Yet if I say "ask Smart heat" and then wait for the skill to invoke before giving the command, it works every time.. I think I will just stick with the latter method which works.

I've found it usually happens in areas of the house where 2 or more Echos pick up the command and likely cannot decide amongst themselves which one of them should process it. In rooms with just a single Echo and the door closed, it never happens (to me). I have an open-plan downstairs with 3 Echos and that's the only place I've seen this behaviour.

[jweaver]

Whichever variant you use, remember to mention an until or duration. Otherwise, it's only until the next scheduled switchpoint reverts things. So if you're coming home from work at 5PM, add "... until 5PM" to your utterance.

Yes.. Cheers.. Thats exactly I am using Evocontrol as the Evohome controllers seriously lacks in that area...

[jweaver]

I've found it usually happens in areas of the house where 2 or more Echos pick up the command and likely cannot decide amongst themselves which one of them should process it. In rooms with just a single Echo and the door closed, it never happens (to me). I have an open-plan downstairs with 3 Echos and that's the only place I've seen this behaviour.

That certainly explains it.. during my testing I can hear several echos waking up.. But I have tried to do it in a room where only ONE can hear me and have no success.. Its no issue.. Breaking it into 2 sentences makes more sense anyway as it gives you time to think.. I am not sure if its just me, but I still find the whole process of 'taking' to a piece of electronics and plastic uncomfortable.. No matter how elequant I am in public, I just find myself a gibbering mess when trying to give Alexa commands.

[philchillbill]

That certainly explains it.. during my testing I can hear several echos waking up.. But I have tried to do it in a room where only ONE can hear me and have no success.. Its no issue.. Breaking it into 2 sentences makes more sense anyway as it gives you time to think.. I am not sure if its just me, but I still find the whole process of 'taking' to a piece of electronics and plastic uncomfortable.. No matter how elequant I am in public, I just find myself a gibbering mess when trying to give Alexa commands.

It's that way for everybody starting out but you quickly get used to it. I control absolutely everything by voice as I have a load of unpublished skills to control older gear for which the original manufacturer has no Alexa support. Even though they're not perfect, the Shows have been a game-changer because a screen makes a world of difference. Try asking EvoControl for a system-summary on an Echo Dot and having to wait for all the zones to be enunciated in sequence vs just glancing at your Show. IMHO, if you look at what an Evohome install itself costs in terms of hardware, buying an Echo Show just for this skill is worth it in terms of the extra comfort/savings possibilities. But I'm likely a wee bit biased

[jweaver]

It's that way for everybody starting out but you quickly get used to it. I control absolutely everything by voice as I have a load of unpublished skills to control older gear for which the original manufacturer has no Alexa support. Even though they're not perfect, the Shows have been a game-changer because a screen makes a world of difference. Try asking EvoControl for a system-summary on an Echo Dot and having to wait for all the zones to be enunciated in sequence vs just glancing at your Show. IMHO, if you look at what an Evohome install itself costs in terms of hardware, buying an Echo Show just for this skill is worth it in terms of the extra comfort/savings possibilities. But I'm likely a wee bit biased

[Midland45]