Control and Query your Tesla vehicles (via Tessie.com)

I would also use that! :blush:

1 Like

bug in time input (from to) has now been fixed in latest non beta release.

4 Likes

When will this latest version be available in Package Manager?

the latest version is in package manager the fleet api changes have not been integrated at this point.

Thanks @kahn-hubitat, is the fleet api change what resolves the time attribute problem? Reason I'm asking is I'm still having problems saving preferences with your driver with the current package manager version. If I comment out the time attributes like @Bloodtick_Jones did than it saves fine. Otherwise nothing happens and the preference page just stays open even if I click "save and close"

as mentioned you need to update the firmware on your hubitat.. what version are you running

Sorry, I missed that part. I was running 2.4.0.150 I just updated to 2.4.0.151 and everything is working perfectly. Thanks again for providing this great driver.

1 Like

New major version available on github and through package manager for testing.

  • v 2.1 many changes. first integration using the new tesla/tessia fleet streaming API. Thanks to Bloodtick_Jones and ALAN_F for initial code stub.
  • Caveats, this is a little different that the non official version:
    1. There is a new input preference flag that needs to be enabled to use the web fleet websocket API. Without this it works just like it used to albeit alternate presence sensing will
  • now NOT work without the websocket API. So all the code and preference flags for reduced refresh are gone and the minimum refresh time is 1 minute. the Reason being is that
  • this methods works much better for frequent data updates to figure out when you are close to home to set whether you are present or not. This is also enabled with a setting.
  • As before alternate present uses your home longitude and latitude and the distance you decide it should fire.. Without all these set it will flag an error in the logs and disable it.
  • Normal presence sensing if you have the homelink module is still better and recommend not using the alternate method if your car came with the homelink module or you bought the add-on (earlier model years it was included for free).
  • There is also a new debug toggle just for the new websocket api, if either this or the older debug_level=full are on websocket debugging will come out. This is so that if issues
  • arrise with the new code you don't have to commit to full debugging of the polling api as well.
  • the following functions and attributes and related input preference have been removed:
  •    input "outerBoundryCircleDistance", 
    
  •    input "outerRefreshTime", 
    
  •    input "refreshOverrideTime",
    
  • FromTime and ToTime settings are still relavent to disable polling and also now will disable the websocket api interface during this time as well. This normally is used
  • when you know the car will not be used ie. late at night to help it sleep and also reduce load on the hub. If the websocket interface is enabled it will also now re-enable
  • when the daily toTime is met.
  • A timeToFullCharge attribute that displays the remaining time to charge (NOTE the name is misleading) it is not really the time to full charge but the time remaining to charge up- to the
  • level you have set. But anywhoo this is what telsa has called this attribute. The attribute is a string similiar to what is found in the tesla app.
  • Other notes: Beyond our control is that the websocket interface seems to reset every 5 minutes, and in order to fill in missing data on intial startup or when this occurs
    (as long as the car is NOT Asleep) a normal polling refresh is fired off. The result is that you will see normal refreshes every 5 minutes if the car is awake.
  • For this reason, normal polling probably should not be set lower than 10 minutes, 15-30 minutes is now the recommended level for the normal API polling.

*/

4 Likes
  • v 2.11 fix typo in disable fx
  • v 2.12 fromtime was not disabling the websocket correctly, the status fx was reopening it.
    */
1 Like

There is a bug with the new textual time to charge attr/code. Finally had charging over an hour.

I debugged and fixed it while at sushi waiting for food
On vpn over phone. Lol

Will post after concert tonight.

1 Like

ok v 2.13 with above fix now in github / package manager

2 Likes

v 2.14 available...

Possibly unnecessary? change to NOT issue a sendEvent update for presence if it didnt change.. i was not flagging it as isStateChange = true so don't think it was triggering a rule anyway,..

But not sure why i saw duplicate events when vehicle was already present.. possibly because presence is both an attribute and a capability.. maybe a bug.

1 Like

solved the issue about not wanting the tesla model 3 without the stalks..

picked up a used 2022 model 3 red dual moter with 19" wheels 9700 miles and accel boost for 31k... direct from tesla... est range was 334 with the 19" wheels should be around 350 with 18" Better than my 2018 with now about 260 with 18" wheels.. sigh

now busy ordering 18" rims, snow tires, winter mats etc.

keeping the other model 3 as well.

1 Like

The 2022 dual motor is the one I have, also with the 19in wheels, and also got the 18in wheels with winter tires.
image

It’s a great vehicle that has very respectable range. Used it twice to go to Clearwater Florida and back from New Brunswick, so about 3000 miles.

It is quieter than the 2019 and earlier models, though probably not as quiet as the new 2024+ model.

Added DIY Detailing’s Ceramic coating on it (was cheaper than getting it done, and there was enough product to do 3 vehicles). It looks great and is super easy to clean now. :blush:

The new Model Y also seems interesting, though is currently quite cost prohibitive…

I went from a 2020 SR+ to a 2025 RWD LR and the difference in ride quality and noise level is night and day. I expect the 2022's are somewhere between the two, as they've been constantly tweaking the suspension and NVH. If they can get the new Model Y suspension to be anything like the new 3, it should be a huge success. Our Model Y is very practical and can swallow a surprising amount of cargo, but I find the ride downright unpleasant most of the time.

The lack of stalks on the Highland has been mostly a non issue for me. Once in a while I find I hit the touch buttons on the steering wheel when I'm trying to use the scroll wheel, but that's more a problem of having ANY controls next to the scroll wheels rather than specifically not having the stalks. I'm still struggling with hitting the right scroll wheel by mistake when I mean to hit the voice command button. I don't drive daily, so I think I'll get it with more time behind the wheel.

The on-screen shifting works just fine as long as you don't have gloves on. I need to take my gloves off to enter the PIN to drive anyway, so it's not a problem. Auto shift works fairly well, so I rarely use the screen to shift A 3-point turn requires just the right driver inputs to trigger auto shift, and I'm still learning that. I always back into my driveway, so every time I return home I should be auto-shifting to reverse. I get it right about 80-90% of the time.

The other new feature that I like is auto shift to park when the driver unbuckles their seatbelt. It completes the automated process of just pulling up to your destination, getting out of the car, and walking away. We already have no 'off' button and no need to manually lock the car, and now there's no need to manually put it in park either. I know the cars already auto park if you take off your seatbelt and un-weight the driver's seat, but that results in a warning/error message.

1 Like

I don't miss the gear selector stalk at all.

But I really miss the turn signal stalk. At least once every trip I seem to errantly signal incorrectly (e.g., left instead of right, or errantly turning on a signal a turn when I don't mean to signal at all).

Basically, I've only been a spastic at turn signals since getting my new car last year.

2 Likes

@kahn-hubitat might it be that there is a bug somewhere in logger level selection? My logs are set to info, but I see an occasional "debug" level message in the logs.

you are correct:

v 2.15 extra debugging in charging timestring removed, and also debugging that was supposed be info and was
coming out related to new alt presence code, changed to only come out for full debugging.
Also removed old code for alt presence that was already commented out.

3 Likes

I assume all is well with the websocket and alt presence changes, as i havent heard anything other than the logging issue?

2 Likes

Seems to be working for me. :blush: