NOAA Weather Plus discussions

[midd]

Just curious if there is a need to have both NOAA plug-ins enabled?

[nsheldon]

That's up to you really. It won't hurt anything to have them both enabled, but it probably isn't necessary. I tried to provide all the data provided by the NOAA Weather plugin in the NOAA Weather Plus plugin. As noted in the main post at the beginning of this thread, NOAA Weather Plus lacks some of the "...String" states provided in the NOAA Weather plugin, mostly because those strings aren't provided in the data feed used by NOAA Weather Plus. The same data is in there though (temperatures, humidity, pressure, wind direction, heat index, etc). Where possible, I used the same state names so people could more easily select NOAA Weather Plus as the plugin source for their existing NOAA Weather devices in Indigo.

[CraigM]

Just installed and got this error:

Code: Select all

Jul 23, 2012 10:25:35 AM
  SQL Logger                      creating table device_history_844513372 for "Weather"
  NOAA Weather Plus Error         File didn't contain an observation date for coordinates 33.727, -116.236 - possible corrupt data file from NOAA

[nsheldon]

Just installed and got this error:

Code: Select all

Jul 23, 2012 10:25:35 AM
  SQL Logger                      creating table device_history_844513372 for "Weather"
  NOAA Weather Plus Error         File didn't contain an observation date for coordinates 33.727, -116.236 - possible corrupt data file from NOAA

Thanks for the info.

I'm able to reproduce this error. Unfortunately, it's an error in the NOAA XML data, not in the plugin. To illustrate this, go to http://forecast.weather.gov/MapClick.ph ... n=-116.236 which is the National Weather Service web page for your location. The NOAA Weather Plus plugin uses the XML version of that page which you can download yourself by clicking on the "XML" image (orange background, white text) on the right side of that page. If you download it and look at the content, you'll see that all the current condition information attributes in that file are blank. I'm not sure what source the HTML version of the page is using to obtain its information, or if it's just using old cached information (you'll notice there's no current condition icon even on the HTML version of the page linked to above), but the source of the XML version doesn't appear to be working properly. The only workaround I can think of is to use a different, nearby location like San Diego, CA (coordinates: 32.71530, -117.156).

[CraigM]

I'm going back to using a nearby airport which does broadcast data.

Great Plugin! Lots of features

Thanks for the effort

[nsheldon]

Thanks! Yea, using a different source location would probably be best for now. NWS just started using a new web interface this month. I wouldn't be surprised if some of the lack of data is at least partially due to their system changes.

[CraigM]

All 5 icons [5 day forecast] on my control page went to 'night' at around 4pm. Is this normal, or is there an adjustment needed?

[nsheldon]

All 5 icons [5 day forecast] on my control page went to 'night' at around 4pm. Is this normal, or is there an adjustment needed?

That sounds about right. As with the current condition information, the icons used are completely controlled by National Weather Service data, so if the icons changed to nighttime icons at around 4:00 PM today, that's because NWS changed their icon URLs for the forecast in your area at most 15 minutes before you saw the change on your NOAA Weather Plus Indigo device.

[nsheldon]

All 5 icons [5 day forecast] on my control page went to 'night' at around 4pm. Is this normal, or is there an adjustment needed?

I was thinking about this more, and I don't think I explained very well why this should be normal. You'll notice that the NOAA Weather Plus device has 15 forecast periods (though sometimes the 15th period doesn't contain any data). These 15 periods are approximately 12-hours apart from each other, relative to the observation time. What this means is that in the morning (sometime after 6:00 AM usually), the first forecast period (the "forecast1..." states in the device) will reflect the morning forecast up to about noon. The "forecast2..." states will be for that afternoon. Around 9:00 AM or so, NOAA will update their observations and forecasts, and the "forecast1..." states will represent the forecast for that afternoon while the "forecast2..." states will represent the forecast for that evening. This shift in representative period takes place every 3 to 6 hours, depending on the local NOAA observation station (I assume). Unfortunately, that does make it difficult to rely on the various forecast states to represent forecasts for a specific time of day. While creating the NOAA Weather Plus plugin, however, it did seem the most consistent and reliable way to keep the number of device states (whose names cannot be changed) as low as possible while still providing all the forecast information.

[CraigM]

I'm getting a lot of errors in my log [± every 20 minutes]

Code: Select all

  NOAA Weather Plus Error         File didn't contain an observation date for coordinates 33.6267, -116.1594 - possible corrupt data file from NOAA

  NOAA Weather Plus Error         Error parsing XML from NOAA for device Weather [NOAA Plus]: not well-formed (invalid token): line 1, column 111

The info on the NOAA website is updating correctly, but on my control page the Current Temp and Current Hum hasn't updated in hours.

FYI,
- The 5 day H/L forecast is correct on my control page.
- The NOAA Weather plugin device is showing the correct data to the NOAA website. Why would the two plugin's pointed at the same weather location display different data?

Any thoughts?

[nsheldon]

I'm getting a lot of errors in my log [± every 20 minutes]

Code: Select all

  NOAA Weather Plus Error         File didn't contain an observation date for coordinates 33.6267, -116.1594 - possible corrupt data file from NOAA

  NOAA Weather Plus Error         Error parsing XML from NOAA for device Weather [NOAA Plus]: not well-formed (invalid token): line 1, column 111

Hi CraigM.

Yea, I'm getting the same error too. I'm looking into it. I think NOAA changed their XML layout.

The NOAA Weather plugin device is showing the correct data to the NOAA website. Why would the two plugin's pointed at the same weather location display different data?

The reason is because the two plugins aren't actually pointing at the same data. While both use data generated by the National Weather Service, the NOAA Weather plugin included with Indigo 5 points to a completely different file (with less information in it) than the NOAA Weather Plus plugin I created.

[CraigM]

Yea, I'm getting the same error too. I'm looking into it. I think NOAA changed their XML layout.

Thanks, keep us posted if there is a fix

[nsheldon]

Okay. I've posted an updated version 1.0.4 here that correctly pulls out the weather information from the NOAA XML even if they don't include an observation time. You'll see an updated error in the Indigo log simply indicating that no observation time was provided and that the data may be out of date, but NOAA Weather Plus will still pull any weather information in the XML out if it's there. Let me know if that fixes things for you (it did for me).

[jay (support)]

Not sure where nsheldon's data feed is getting it's data, but our NOAA Weather plugin gets its data from the National Hourly Aviation Weather Observations data feed. It was one of the first feeds they made available in an RSS feed. It's been very reliable the last couple of years (less so in years before that), but doesn't have a lot of data in it - no precipitation because apparently airports don't really care about rain (according to the guy at NOAA that I corresponded with years ago).

[nsheldon]

The feed I use is taken from one of their newer DWML feeds (http://www.nws.noaa.gov/mdl/XML/Design/ ... Design.pdf. So far is hasn't been as reliable as the Hourly Aviation Observation Data, but, as Jay says, it has more information in it. It also provides slightly more granular observation data than might appear in the aviation observation feed. Trade-offs I guess. Features for reliability... at least until NOAA gets the DWML generation to an equally reliable state.