Fix it Friday has advantages of getting ongoing weekly reports on an issue that the most people care about. They can’t afford to give detailed action plans or weekly reports about every single thing they are working on though. Fix-it-Friday bugs are not the only bugs they fix or address. Many such things are worked on and fixed constantly, including behind the scenes. I often pass on issues and logs to some devs I have contact with or through Jason, or others and I see a ton of things get worked on and fixed all the time that are never listed in Fix-it-Friday. The main reason I love to report things in fix-it-friday is not because it is the only thing Wyze does something about, but because I love to have a weekly update/report on how things are going for the issues that affecting the largest amount of users. That’s the main point of “winning” fix-it-friday…not that it’s the only way something gets addressed, but that users get to decide what we most want the most details on. Most companies don’t give out any info or share with users what is going on at all. You just wait for them to tell you when a fix is ready, and that’s the only communication you ever get. Wyze allows us to decide what issues we want to hear more about. That is rare thing for a company to do…and I say this as someone who has worked professionally as a QA tester with multiple clients doing software testing, etc.
But I can totally see how it can appear that way (they only look at an issue if it wins the vote).
Honestly, it will be a sad day if Wyze ever gets rid of Fix it Friday because of this misperception. We will get a lot less transparency and not be able to have as much influence in telling them what is the most important thing to users. We’d lose a huge voice, and they’d become more like every other business that just doesn’t share any of that info. That would be a sad day.
Dang, sorry to hear it didn’t work. I’d experiment some more, but it sounds like i better not touch anything (and not upgrade firmware) now that i have it working for my use case.
I just bought an Orbit B-hyve and am trashing the Wyze unit officially today, after more than a year of dead grass, dead plants, support frustration, and headaches.
Good luck with everything, everyone. Hope they get it fixed for you all.
I live in Texas where we are experiencing a heat wave with temps near or over 100 for at least the next 14 days. There is no rain or wind to speak of. There is obviously not going to be a freeze. I have a smart schedule to run on Tuesdays and Fridays (the only days my city allows). Today, despite the system saying saturation is below 10%, the system didn’t run and isn’t in the schedule. It is next set to run on Friday, with water saturation at 0 until then. I set a manual schedule for today. Now it is no longer scheduled to water this Friday either, despite it predicting saturation of 13% and 0% on Saturday, Sunday, and Monday. Why?
I continue to have problems with the skips on the latest firmware. Even with less than 3 zones. Just woke up to check what it would do since it’s supposed to run soon and it says the sprinkler unit is offline. I’m also in Texas and this sprinkler did end up killing my grass (all of it in the backyard, most of it elsewhere) from lack of watering.
Seeing it offline does mean I get to see that the sprinkler will wait 2 weeks before doing the offline schedule so that explains why nothing happens even when it doesn’t run. If we could remove this wait that might help most of us. Like I said, I’m in Texas where it’s probably not going to freeze anytime soon so the 2 week buffer for me to change anything isn’t needed (assuming that’s what it’s for)
I will try to set up my replacement this weekend but how sure are we that this isn’t a cloud problem?
I assume that the schedules are ultimately saved in the cloud and triggered by the cloud with the sprinkler unit simply following instructions. Could it be that there are redundant/failover servers that are not synced properly?
I’ve mentioned it before but mentioning again the time that I watched it not do the previously scheduled watering (went minutes past the start time yet continued to say it was up next). That day I changed the schedule forward a few minutes and then it worked despite just being skipped moments before. This would go with the cloud issue theory: the original schedule was on Server 1 but the failover, Server 2, was active at the time the schedule was supposed to run. Server 2 did not sync the schedule so it didn’t run. Moving the time a few minutes saved it to Server 2 which would now have a schedule and run it. The next week when it’s supposed to run, Server 1 is active again which also didn’t sync and no longer has a schedule. Therefore it skips for without reporting since it didn’t have anything scheduled. The cloud logs would also show nothing wrong because it didn’t have a schedule.
Of course, I have no idea what Wyze’s architecture is like and this is an attempt to oversimplify for the sake of explanation but I figure I’d put it out there. I am a little confused by the fix it Friday update saying the cloud logs look normal since wouldn’t the logs be missing the entire schedule? That doesn’t sound normal
It just skipped 4 AM schedule. After it disappeared from “Next”, I moved it to 4:15 AM to try again. It added it back to “Next” but proceeded to skip it again. I then moved it to 4:17 AM before it disappeared from “Next” and this time it ran.
Logs:
Before 4:17: 1091957
After 4:17: 1091960
The only similarity with last time I did this and it working was that the skipped schedule still appeared in “Next” when I adjusted it
Manually triggering runs from the app work successfully
Questions/Guesses:
I assume that a manual call to turn on a zone initiates a call from my phone to the Wyze cloud. How is this message communicated to the controller? Either a push from the Wyze cloud to the device or a frequent poll from the device to the Wyze cloud. In my case, I’m behind carrier NAT and a typical home firewall, making server-to-client initiated more difficult. @WyzeJasonJ , can you confirm how communication between the controller and wyze takes place
As implied by my messages, I agree this is a cloud issue. Same experience (no smart to time based schedules, trigger from app works every time for the past year up until 1 failure last week).
I do like that we were added to fix it Friday so hopefully that gets more eyes on it. My concern for the last update is that they said the logs look normal. If it’s missing a schedule in the logs, I wouldn’t consider that normal. If the logs say they triggered a schedule and we’re saying they didn’t (and it’s not in the history), I would also say that’s not normal
I have a partial theory that doesn’t entirely explain everyone’s issues, but it’s tangentially related, so please hear me out.
I think there is an issue with the cloud service / smart schedule logic. I suspect that there is a bug on how it determines whether or not a zone needs to be watered. Consider the situation where it has previously scheduled Zone X to water tomorrow at 8am. Sometime between now and tomorrow the cloud service determines it doesn’t need watering (maybe because it thinks it rained or it thinks it is going to rain). It revises the schedule so it will not water tomorrow, but instead in 3 days. Tomorrow 8am rolls around and Zone X doesn’t water. There’s no notification that it skipped, because in it’s own electronic mind, it didn’t skip it - it wasn’t scheduled anymore so it did exactly what it was told to do. The logs would look correct because the system says it wasn’t supposed to run and it didn’t run.
If this is the case, this is particularly frustrating to those of us in the South. In the summer months, there is a slim chance of rain/thunderstorms nearly every day. But most days, there is no precipitation at all. No matter how I set the skip thresholds, the smart schedule keeps moving the watering days to later (in my app at least, I cannot deselect rain or saturation skipping for smart schedules). Even though it hasn’t ACTUALLY rained in many, many, many days and the temp has been over 100 for a long time. I’ve been manually watering with my garden hose and setting manual schedules, which kind of negates the use of a smart controller and the $20/yr fee.
My suggestion: enable a verbose notification mode for debugging. That is, every time a smart schedule is dynamically changed or recalculated, let us know what change was made and why.
