WeatherSnoop Problems - constant trigger and disabling

[JGTracy]

I've come across a couple of problems with the WeatherSnoop plugin: One problem is with the "any change" to "rainTodayInches"; and the other that the plugin can suddenly disconnect from Indigo.

The "any change" trigger for "rainTodayInches" triggers constantly if there is no rain at all. It only works as expected on days that have at least 0.01 inches. (My Davis Vue records rain in 0.01 increments.) This seems to have an easy work-around by including the condition that there must be at least 0.005 inches to execute. Easy work-around or not, it should be fixed.

The other is more troublesome. The "Plugin "WeatherSnoop" disconnected" is recorded in the log, but I cant find a way to create a trigger from that so it can be reloaded. I don't know if I could create an action to reload the plugin even if I could get a trigger. "Reset all" might work, but I won't be able to test that till it disconnects again.

Seems to be working fine for my needs otherwise.

FWIW WeatherSnoop (1.0.1) and Indigo (5.0.2) are running on the same admin account on the same iMac.

[jay (support)]

Hmmm. I just created a trigger that logs whenever there's a change to that state and it's yet to fire (and it hasn't rained today).

When you select the weather station, what shows in the state list for rainTodayInches?

[jay (support)]

Oh, and the next release will fix the periodic plugin disconnect.

[jay (support)]

Also, make sure you have the rainTodayInches state as the trigger - I found a bug on other rain states (but not that one) that might cause what you're seeing. It will be fixed in the next release as well.

[JGTracy]

Thanks for the info on the periodic plugin disconnect.

As for the rainTodayInches, my trigger is working fine today also without the 0.005 inches condition. This may be because it rained yesterday and WS inserts a 0.00 in the data to "zero" things at midnight on the day after a rain. (The WS data custodian shows this happened at 00:00:04.) This notion seems to be supported when I checked back through the logs. If it doesn't rain today, I should be able to confirm this tomorrow. Note, however, that I will be away from my computer for most of the rest of the year and may not have the time to respond before I leave.

[JGTracy]

The trigger is as follows:
Type: Device State Changed
Device: DavisWeather
Rain Today Inches
Has Any Change

I presume this is what you mean.

[jay (support)]

Yep. I did find a different but (which shouldn't be effecting the rainTodayInches state). Keep me posted as you can - I'll also leave my trigger running (supposed to rain tomorrow but not today) so over the next couple of days hopefully I'll see it fire incorrectly.

[JGTracy]

As expected: rainTodayInches triggers constantly if there is no rain today and was no rain yesterday. Here is a small part of this morning's log:

Dec 24, 2011 12:00:01 AM
Deleted old log file "2011-11-24 Events.txt"
Created new log file (Indigo Server version 5.0.2)

Dec 24, 2011 12:00:37 AM
Trigger RainTest
Trigger RainTest

Dec 24, 2011 12:01:37 AM
Trigger RainTest
Trigger RainTest

Dec 24, 2011 12:02:37 AM
Trigger RainTest
Trigger RainTest

The trigger is occurring at a 1 minute rate. In this test, it modifies a variable first to true and then to false. I expect the same thing could happen when "Any Change" is used with other rain variables such as rainMonth, rainYear, etc, but less likely to be a problem.

Have a happy holiday.

[jay (support)]

Yep, mine is too. I'll have to dig into this more next week.

[johnpolasek]

Mine also triggers once a minute on any change in rain rate... although the variable posted remains 0.

[jay (support)]

Probably won't get a chance to look at it this week but it'll be first thing on my list next week.

[johnpolasek]

It's not really an issue for me; all it seems to do is clutter the log, since I DON'T have the shutdown problem and the variables the triggers are posting to remain 0... I was just making sure you knew that it affected multiple items.

[JGTracy]

This problem may have been fixed by WeatherSnoop. I just got a notice of an update (2.0.9) that (among other things) fixed a problem with zeroing some variables. I downloaded and re-enabled my rain test trigger and have no triggers. -- yet

[jay (support)]

Interesting...

[JGTracy]

The problem is not fixed, although it seems to be partially fixed. After several days of no false triggers for changes in rain today when there has been no rain, The trigger restarted this morning after midnight. I think WeatherSnoop was on the right track in fixing variables that were not properly zeroed but may have missed some. -- Like month rain and year rain.