Epix (gen 2) with 13.22 - onEnterSleep fires before watch sleeps

Has anyone else noticed an unusual quirk with the latest epix software and the onEnterSleep function timing?  I noticed this myself, and I've had some customers write in asking for help.

For ConnectIQ watch faces (mine and other developers'), the onEnterSleep() function seems to be getting called a few seconds before the device actually sleeps.  This causes the low-power mode version of my faces to display before they're supposed to.

Video of the behavior (from a customer) is here:
https://photos.app.goo.gl/cg3TQPYRGtECi3Ve8

This customer has AOD off.  Around the 8-second mark, my watch face goes into the low-power mode.  But the watch face doesn't actually turn off for another few seconds.

This seems to be impacting all my faces now, and it's also impacting other developers' faces I downloaded from the store.  All mine were all built with 4.2.4, and they were working fine until customers updated to 13.22.  

Notably, built-in Garmin faces are not having this issue, so it seems related to Connect IQ only.

Thanks!

Mike 

  • With AOD off. under settings>System>display>general usage, what is the timeout set to?

  • That customer has a 15-second timeout.  I also have a 15-second timeout.    

    Here's another video showing what's happening with MobileDriveway's Glance:

    https://www.dropbox.com/s/nw2z4bkz18alsct/IMG_2134.MOV?dl=0

    In this video:

    • When I go to wake up the watch with the "light" button, the screen turns on and the watch face exits sleep mode, but instantly the watch face returns to "low power" mode.  (I turned on the beeps with the buttons to make sure I wasn't double-tapping the button)
    • I press the "light" button again at the "03" second mark, and it stays in high-power mode until "13".  Then the watch face goes into sleep mode (after only 10 seconds).  5 second later, the watch screen itself goes into sleep mode.  
  • Looks to me that things are pretty much happening as they should.  I don't have an epix2, but do have a 965.

    After a gesture, a CIQ WF drops to low power (onEnterSleep) after 10 seconds.  This is NOT based on the timeout setting I mentioned).  So if you have the timeout set for 15 seconds, do a gesture to get a WF out of low power, after 10 seconds the WF goes back to low power mode, and after 15, the display turns off.  Native WFs have a whole different set of rules.and onEnetrSleep()/onExitSleep() are only in CIQ WFs.  On the 965, pressing the light button acts the same as a gesture.

  • Thanks, Jim.

    This behavior is new to me and my customers reporting issues. I don’t believe this is working as intended, unless the spec just changed  

    Do you have a reference to where this 10 second timeout is documented?

    Because until this update, all watch faces stayed in high power mode until the screen timed out. They didn’t used to go to sleep at different times. If the screen timeout was 15 seconds, the watch face stayed in high power mode for 15 seconds. 

    Also, note how in the video the watch face wakes up right into its low power mode. That’s definitely not right. 

  • It's been 10 sec for dropping back to low power since day 1 of CIQ and it might be documented some place.

    All I can say is on the 965, things work as I expect them to work.

    In the videos, looks the same to me.  After 10 seconds, CIQ drops to low power, shows the low power screen for 5 seconds then goes blank.  If you reduce the general usage timeout to 8 seconds, you won't even see this.

  • The power and light behavior of AMOLED devices is different from MIPS devices.

    I noticed the same thing, and for AMOLED devices this issue was not present before 13.22.

    If the backlight timeout is set to 15 seconds:

    before 13.22, after 15 seconds with no input onEnterSleep is called just as the backlight gets dimmer simultaneously.

    And after onEnterSleep, any input invoke onExitSleep with backlight on.

    But on 13.22 after 10 seconds of high power mode, even there were some key (back key) and/or touch input, onEnterSleep is called.

    Funny thing is that the backlight is on for 15 seconds after the key/touch input.

    While the backlight is on after onEnterSleep is called, pressing LIGHT key make the device to go 'real' low power mode with backlight off.

    This makes kind of weird user experience - when user saw the watchface goes low power mode after 10 seconds without input(onEnterSleep called and watchface-wise low power mode) and the user want the watchface to go high power mode, the user should hit LIGHT key twice. (first press - backlight off and device-wise low power mode, second press - backlight on and onExitSleep called).

  • Hi Jim, your experience may be true on other devices, but I'm not sure that's true on the epix.  Have you had an epix gen 2 to comment about?

    I've used the 15-second timeout while recording demo videos of my watch faces, and they consistently used to sleep at exactly the same time the face did.  There was no different 10-second timeout for the face vs. 15 for the screen.  I especially noticed this problem while trying to record a video for a new face I just built & submitted.  The face kept going into low-power mode sooner than anticipated.  And on top of that, tapping the screen stopped waking it, now, too

    I'll also note that other customers in the main Epix (gen 2) forum are reporting screen-related weirdness with 13.22, too, so this isn't something to just dismiss as "always worked this way".

    This seems related:
    https://forums.garmin.com/outdoor-recreation/outdoor-recreation/f/epix-2/335630/random-wrist-gesture-all-b-w-display

    I believe Garmin was trying to resolve this issue and introduced another:
    forums.garmin.com/.../watch-doesn-t-always-go-back-to-always-on-display

  • Yes, , that's exactly what I'm experiencing.  This is a pretty unusual behavior.  Thanks for seconding.

    This part is particularly frustrating, since as a user I can't keep the watch face on high-power mode if I'm reading some of the data longer than 10 seconds:

    But on 13.22 after 10 seconds of high power mode, even there were some key (back key) and/or touch input, onEnterSleep is called.

  • Confirming Garmin acknowledged the bug, reproduced it, and is making a ticket to resolve it (some of this context via PM from Garmin - Chris):
    forums.garmin.com/.../watch-faces-go-to-sleep-too-quickly-buggy-after-13-22

  • Hi,

    Well, I've just received a video from a user which seems to be a similar issue, it seems the onEnterSleep is not trigerred in a right time.

    (I don't have yet his autorization to post it)

    The user is on AOD off

    you can see on the video that the screen turns off (as expected) then sometime when he touch the screen the WF awake for less than 1 sec on goes to my "simple display" which is triggered onEnterSleep