Tutorial: Single scene instance / Event model

[jgab 2,642]

12 hours ago, domin12 said:

ok, will test it after upgrade, i'm still using old version _version,_fix = "1.15","fix12"

o my..... so many things was happend here , how to quickly upgrade my old version ?

trying move to new version, but after moving my rules got this error :

ps. using HomeTable variable

 There should be no changes to your rules so the upgrade is not that difficult.

It seems like you don't get to initialise your hometable and it becomes nil

 

The hometable setup in beginning of EventRunner.lua is bit different.

The variable pointing to your FibaroSceneAPI hometable variable is named '_homeTable' now...

...and you need to uncomment the line "local HT = type..."  in the beginning of your main()

Please login or register to see this code.

..and then HT is used for the table (you can delete the inline HT example).

Note that I call Util.defvars and Util.reverseMapDef on HT.dev as I have the device id's in the .dev "branch" of the hometable. There could be other branches in the hometable that is not nodes ending with device numbers (like the "other":"other" in the example) and calling defvars/reverseMapDef may cause some confusion.

You are of course free to change it so it fits your hometable structure. 

Otherwise it should be no changes - except if you run the emulator and talk to node-red - then you need to update the node-red flow a bit.

Let me know how it goes 

 /J

 

 

Edited March 15, 2019 by jgab

[jgab 2,642]

23 hours ago, petrkl12 said:

it looks that it will be really clever :) without dependency on Fibaro:setGlobal and getGlobal

 

I see one potential problem and it's a automatic delete of log ...

 

When it resets the log (every ~250 entry), it sets it back to the last 3-7 entries - it varies. Unfortunately, that means that the consumer can lose events if it doesn't read them fast enough... experiments seems give 1-2 lost events at every turnover if producer is twice as fast as consumer.... so back to square one.

[domin12 9]

4 godziny temu, jgab napisał:

The hometable setup in beginning of EventRunner.lua is bit different.

The variable pointing to your FibaroSceneAPI hometable variable is named '_homeTable' now...

...and you need to uncomment the line "local HT = type..."  in the beginning of your main()

Please login or register to see this code.

i have change it before, so the error is still the same 

Please login or register to see this code.

Does this mean that my HomeTable structure is wrong?

[jgab 2,642]

4 minutes ago, domin12 said:

 

Does this mean that my HomeTable structure is wrong?

Do you have a ’dev’ key at the top level of your home table with all your devices?

otherwise change defvars and reversemapdef to only get HT as argument and not HT.dev etc

[domin12 9]

10 minut temu, jgab napisał:

Do you have a ’dev’ key at the top level of your home table with all your devices?

otherwise change defvars and reversemapdef to only get HT as argument and not HT.dev etc

Exactly at the same moment in which you wrote the message. I understood what I did wrong, it looks like everything will be fine

 

--update 

Works great

Please login or register to see this image.

/monthly_2019_03/image.png.189f343e1d6b663e9e5f2a4f9c8f007a.png" alt="image.png.189f343e1d6b663e9e5f2a4f9c8f007a.png" />

 

how to put the percent symbol to this rule after moisture value 

Please login or register to see this code.

I'm trying 

Please login or register to see this code.

but as You probably know, it does not work.

Edited March 15, 2019 by domin12

[jgab 2,642]

18 minutes ago, domin12 said:

how to put the percent symbol to this rule after moisture value 

 

You need a %% to print a % with string.format

Please login or register to see this code.

 

[domin12 9]

8 minut temu, jgab napisał:

 

You need a %% to print a % with string.format

Please login or register to see this code.

 

Thank You

[jgab 2,642]

42 minutes ago, domin12 said:

Thank You

string.format has a lot of useful formatting directives. 

Please login or register to see this code.

gives "%3.14"

and you can do left and right adjustments with padding etc.

Please login or register to see this link.

 

[petrkl12 228]

11 hours ago, jgab said:

 

When it resets the log (every ~250 entry), it sets it back to the last 3-7 entries - it varies. Unfortunately, that means that the consumer can lose events if it doesn't read them fast enough... experiments seems give 1-2 lost events at every turnover if producer is twice as fast as consumer.... so back to square one.

 

it's a pity

 

or you could clear log yourselves and setup some index that you will do it before .... but it will be maybe too slow ...

Edited March 15, 2019 by petrkl12

[jompa68 121]

@jgab

can you see why this is not working? Timer function does not work when using #VirtualSensor

Please login or register to see this code.

 

[jompa68 121]

On 3/13/2019 at 2:30 PM, jgab said:

Is this still a problem? I have tested and have no problem on the HC2.  If I in the emulator set "Hc2.remoteDevices(713)" it works from the emulator too.

On the HC2 does its log the "fibaro:call(713,"setProperty",...)" ?

I can see that VD slider popups to the value i set but then it goes back to what it was before.
Looks ok in debug window (setProperty)

 

When i set value with virtual sonos ID it works.

Please login or register to see this code.

 

[jgab 2,642]

55 minutes ago, jompa68 said:

I can see that VD slider popups to the value i set but then it goes back to what it was before.
Looks ok in debug window (setProperty)

When i set value with virtual sonos ID it works.

Please login or register to see this code.

 

 

Do you have some VD loop that sets it back? If you don't see any fibaro:* log in ER it must be someone else that sets the value? Some other scene?

[jgab 2,642]

Please login or register to see this code.

You have the right scene headers? power in the first and value in the second?

I would make the test >0.5 in the first scene before deciding to publish - that way the second scene get fewer events.

[petrkl12 228]

@jgab

 

why this condition is also true for value=0 ?

Please login or register to see this code.

 

Edited March 17, 2019 by petrkl12

[jgab 2,642]

36 minutes ago, petrkl12 said:

@jgab

 

why this condition is also true for value=0 ?

Please login or register to see this code.

 

Both rules trigger on 100:value, 101:value and 102:value

The second rule is only true if the value (of the device that did the trigger) is equal to 1

The first rule assigns 1 to the value property of 100,101,102 -- which is always true, and thus the right hand side will execute.

[petrkl12 228]

Thanks, but it looks that second rule doesn't work and I don't know why ...

[jgab 2,642]

27 minutes ago, petrkl12 said:

Thanks, but it looks that second rule doesn't work and I don't know why ...

Oh, {100,101,102}:value is the same as {100:value,101:value,102:value} and that is not equal to '1'

To get the id that triggered the event do

Please login or register to see this code.

:dID selects the device id from the table that caused the rule to trigger and returns it, so that :value is applied to the right trigger

[petrkl12 228]

It works . Thanks

I think there are a lot of new rules that could be perfect to describe in wiki.

Edited March 17, 2019 by petrkl12

[jompa68 121]

still fighting with this...

 

Please login or register to see this code.

 

Please login or register to see this attachment.

[jgab 2,642]

I assume the rule runs every weekday between 5 and7.30. Do you get the error every day or once a week or?

I haven't been able to replicate it with a similar rule, so let me think about a better debug statement to capture the bug next time it happens.