Different oberservation times in WW and on Weather.com

No, I have enough on my plate. Do you want to have a go at it?

Won’t superman75 need to know which server WW is using to pull the weather.com data from when he talks to the weather.com people?? I’m not sure how receptive they will be to a report that the WW forecast data doesn’t match their forecasts. But their desktopfw data out of sync with their web forecasts might get some attention.

Is this what WW is using?

http://desktopfw.weather.com/weather/lo … d=3&hbhf=8

Or this one?

http://xoap.weather.com/weather/local/U … d=3&hbhf=8

EdP, all good questions, but I think you’ve missed the most important point. How can the problem be reproduced? That one is a mystery in itself. I don’t think they are even going to blink an eye if someone contacts them with an issue that cannot be easily produced.

I’ll run some automated testing today and see if I can document the problem.

Who says they can not reproduce it?

If superman75, and or others, documents the problem using screen prints of the data feed and the weather.com webpage repeatedly and sends it to TWC it should get their attention since it’s all their data.

If the problem is reported to TWC as a discrepancy between weather.com and Weather Watcher its almost quaranteed they will consider it a WW problem and won’t look into it.

With the testing I did today, I found the XOAP feed had an incorrect observed time at 12:04 PM today. The following XML document from their feed includes the bad data.


Now, who wants to volunteer to send it to them? :cool:

Where is the corresponding weather.com forecast image showing the discrepancy?

And to prove that this isn’t a fluke occurrence there will need to be more than a single recorded instance.

EdP, I’ll leave that detailed investigation up to you. Good luck :thumbright:

:iconbiggrin: You can’t afford me. :razz:

But superman75, the OP, documents things rather well.

What I suggest to whomever decides to report the problem is they present several IE windows showing side by side the weather forecast on weather.com and a corresponding window showing the pertinent XOAP data next to it and highlight the descrepancy. I also suggest leaving out any specific reference to WW.

It would be interesting to know if the weather.com desktop app has the same problem. If so by all means include screen prints of it’s forecasts.

On the other hand, since we now know that the problem is data related, shouldn’t it be easier to fix WW to not display the older data? Maybe when testing the date/time stamp of the new data with what is stored in the Registry, if the new data is less current, temporarily modify the time for the next WW data retrieval to be 2 mins into the future regardless of what frequency has been set.

Perhaps. Is the data old… or the time stamp just incorrect? I’ll have to look through the data I downloaded today when I have some free time.

I am not a programmer so I am sure I have no idea what I am talking about. :iconbiggrin:

But I still really think this Observed time issue is something that would need to be fixed in Weather Watcher.

When I go to weather.com their Observed time is always correct. Its Weather Watcher that is wrong.

If we did report this to weather.com I bet anything they would say the same thing. :cry:

Also like I think I said in an earlier post when you released the Weather Watcher build that fixed the issue where Weather Watcher was getting old data thats when this Observed time issue started.

So it seems by fixing one thing it broke another.

I have a similar problem with WW 5.6.26 on this PC at work - the Observed time is showing as 30/08/2004 in the main window! The Downloaded time is correctly showing as 30/04/2008. I assume it must be something to do with the date format?

In the registry the Downloaded value is 30/04/2008 12:38:35 (UK format, which is where I am!) and the Updated value is 4/30/08 12:20 PM Local Time (US format).

Could this be fixed please, if it is indeed a WW bug? It was doing the same yesterday too :sad:


CiaraJansonFan, yes, it’s a formatting issue. I’m working on a fix.

OK, thanks!