Cumulus Support

Version 1.8IH uploaded to first post in this thread.
Change: handling of communication errors improved - maybe even fixed thanks to Beteljuice

.... and before the glamour wears off, you are respectfully reminded that the weather graphics are FORECAST conditions, not current weather

3 gadget running....normally 20% opacity.
These sites are within about 5 km but lots of difference in the forecast.
Needs a Label to to differentiate the sites.

Needs a Label to to differentiate the sites

But Terry, you're only supposed to be stealing your own data !

Besides, where would "JANTER Electronics Ltd, Weather" fit

I'll have to make a shortcut and park it over the gadget

Version 1.9IH available for download from first post of this thread.

Note: version 1.9 is named cumulus19, so you can install it without losing your version 1.8 which is called cumulusih.
(Both versions are available for download. Make sure you select the one you want.)

Changes in Version 1.9IH:
- 'Last updated' text string above time stamp at the bottom
- temperature and pressure trend arrows preceding numbers
- 2nd temperature field: may be set to any of realtime.txt's fields - or left out
(default: field 24: wind chill)
This variable is NOT on the settings screen, only in settings.ini
- settings.ini: js language: variable names/values, comments
Make yourself familiar with the contents and format of the new settings.ini!

If you don't like wind chill as the 2nd temperature field after the slash, you may choose dew point, heat index, tendency, or any field from the realtime.txt (even Cumulus version...) - or leave it out altogether. That's what the new settings.ini parameter is for.

All working OK Dane.
Perhaps the "Last Updated" could have become the realtime ID? ..... i.e. "JANTER" Just a thought.

I hadn't thought that the use of the realtime.txt from somebody else's site would be theft? Ah well...too bad

Diolch! (Thank you in Welsh)

The work you are putting into this gadget is much appreciated! It is developing very very nicely.

Steve

Hi,

Just been reading through, and I must say it is great to see how the gadget has developed since my early releases.

If it hasn't already been figured out, what is shown below that looks like 3 attempts is actually just the XML reader going through its various states.

1 - LOADING:The load is in progress. Reading persisted properties, but not yet parsing data.
2 - LOADED: Reading of the persisted properties completed. Reading and parsing data, but the object model is not yet available.
3 - INTERACTIVE: Some data has been read and parsed, and the object model is now available on the partially retrieved data set. Although the object model is available during this state, it is read-only.
4 - COMPLETED: The document has been completely loaded, successfully or unsuccessfully.
(http://msdn.microsoft.com/en-us/library ... S.85).aspx)

TNETWeather wrote:Looks like it is making multiple attempts (very fast ones which might be the problem)

Code: Select all

12 January 2010 17:41:24: Enter readyStateHandler
12 January 2010 17:41:24: xmlHttpObj.readyState = 2
12 January 2010 17:41:24: xmlHttpObj.status = 404

12 January 2010 17:41:24: Enter readyStateHandler
12 January 2010 17:41:24: xmlHttpObj.readyState = 3
12 January 2010 17:41:24: xmlHttpObj.status = 404

12 January 2010 17:41:24: Enter readyStateHandler
12 January 2010 17:41:24: xmlHttpObj.readyState = 4
12 January 2010 17:41:24: xmlHttpObj.status = 404

3 attempts for the same file in the same second?

I would think that anyone decent with JavaScript would be able to figure it out. There is a lot of code out there that does the same type of function and its all plain text source.

Well done to dane, for taking the gadget on and if you can't work out some of the code, drop me a PM and i'll see if I can decipher it!!

Steve
In reference to the forecasts as shown on the 3 gadgets i have running temporarily.
There seems to be a problem with the forecast being wildly inaccurate as shown by the first and last gadget, both are fine offset and the centre one is a davis.
Over the last week we have had fairly settled weather, no rain and lots of sun but the forecast has been nowhere near what it is currently.
It may be that the formula doesn't work too readily for Auckland as it is very changeable.
Any thoughts on how to improve the forecasting? Pressure is set to 971 and 1036 in my case.

The forecasts are actually 1 level too bad, and the "No Forecast" was given the "Fine" graphic.
Ergo forecast is artifically bad and the Davis installation has Cumulus forecast turned 'off' so displays the fine graphic instead of a 'blank'

ib has the fixed code in hand.

Website and C1 screen are showing "Unsettled, some precipitation" when the gadget is as the pic. Suggests the error is coming from C1?

As I understand it NZ is "Unsettled, some precipitation" and in some areas a lot of precip atm !

Don't forget there are a limited number of graphics which have to show some subtly different forecasts.

Not in Auckland....we are having a drought. South Island is copping the bad weather.

Super-T wrote:Steve
In reference to the forecasts as shown on the 3 gadgets i have running temporarily.
There seems to be a problem with the forecast being wildly inaccurate as shown by the first and last gadget, both are fine offset and the centre one is a davis.
Over the last week we have had fairly settled weather, no rain and lots of sun but the forecast has been nowhere near what it is currently.
It may be that the formula doesn't work too readily for Auckland as it is very changeable.
Any thoughts on how to improve the forecasting? Pressure is set to 971 and 1036 in my case.

Yes, despite lowish pressures (1000-1008 hPa) the last several days, weather has actually been reasonably settled. Those pressure values usually correlate with unsettled conditions here. I suspect its the pressure that is throwing the forecasts off...my Davis forecast has also been incorrect for several days, quite often mentioning things like precipitation likely, possibly heavy, windy.