Widget: Weather

Former Member
Former Member
https://apps.garmin.com/en-US/apps/7a83e98d-6df2-4920-a90c-9da7275c41f9


Quickly check the weather before you go out for a run. Your last known GPS location is used when you are indoors. This widget requires that Garmin Connect Mobile is running on your cell phone so that it can communicate with the internet. This apps uses the System-Units-Distance setting to determine whether to show statute (Fahrenheit/mph) or metric settings(Celsius/kph). You must be running the latest version of the device software that supports widgets.
  • Brian, unless there is a way to see the bogus value returned on the 920, can be seen as "use distanceUnits", and other wise, use the return from temperatueUnits. What's bogus with the return? What value? Is it not "metric or statue"?
  • BTW I can't believe how outdated the data from "api.openweathermap.org/data/2.5/weather" is...

    I threw together an app (so there would be no revert timeout issues on the vivoactive...), and right now, it's showing a temp of 57, when it's actually about 80 (i hard coded the lat/lon so it's that of where I live, and the location name returned from OW proves it's correct). There is no timestamp with the data, but really, how OLD is this data? it hasn't been 57 since sunrise, and it's now sunset!
  • Former Member
    Former Member over 10 years ago
    The problem is that the temperatureUnits doesn't return a bogus value. It probably just always returns the same value. The data availability isn't checked in the system when it is fetched, so the value in the Monkey C object isn't nulled out like it is for other data values that aren't available (in other APIs).

    When I was working on the weather sample I found that the open weather map API usually returned accurate temperature data, but I did see it return 70 degrees one day in February when it was about 25 degrees out and probably hadn't been 70 in months. It also has a habit of returning Morse, KS for our location, which isn't really an actual place anymore.
  • Ran more tests today with open weather, and for my location (a large city), the temp is off, and temp_min and temp_max are always the same as temp. (I'm assuming that is meant to be low temp and high temp). Same if I call open weather with the city name vs GPS. It could vary by location based on what station OW is getting it's data from, and how often that data is updated by the station.

    update a few hours later: using q=chicago, I've been getting ow data for chicago every 15 minutes or so, and it seems that it's only updated every 2-3 hours at best on the ow website.
  • Brian, does the std weather widget on the vivoactive also use open weather as it's source? If so, I can see why people are complaining in the vivoactive forum about how bad the data is! I mean getting data from a website with 2-3 hr old data at best, just won't work if you want to see something fairly current, no matter how fast that data is displayed or where it's cached!
  • Hi love the widget. Would be good to be able to chose F/C as in UK many of us use MPH and Celsius
  • Former Member
    Former Member over 10 years ago
    Hi love the widget. Would be good to be able to chose F/C as in UK many of us use MPH and Celsius

    Thanks for the suggestion, I will work on adding a way to override the default settings from within the app when I get a chance.
  • BTW I can't believe how outdated the data from "api.openweathermap.org/data/2.5/weather" is-...

    I did a _lot_ of experimenting with OpenWeatherMap in the past and did not see this problem. I gave up on writing it because the makeJsonRequest() method was so unreliable.

    There is no timestamp with the data, but really, how OLD is this data?

    There are timestamps for all weather data coming from OpenWeatherMap.

    If you are just doing a current weather lookup by city name or id, it includes a unix timestamp in utc time. Here is a sample query for Corvallis, OR by name and by city id. Note the dt value in the json response.

    the temp is off, and temp_min and temp_max are always the same as temp

    You are misinterpreting the data. When looking at current weather data, temp_min and temp_max are the minimum and maximum temperature values reported at that time (the time described by the timestamp mentioned above). i.e., with multiple weather stations scattered around an area, you will get variance in measurements. These values represent the extremes of the measured range. If you are looking at a response to a forecast query, the min and max values are what you are looking for.

    Travis
  • Thanks Travis. I'd missed some of the options. I was just using the call as defined in the weather sample!
  • Former Member
    Former Member over 10 years ago
    Hi. Nice App. But I do see wrong data. It is storming a bit with winds around 40 kmh, but the app is showing 9 kmh.
    When I can see timestamps and location info I can judge the data much better.
    Cheers