Cumulus Support

and Control Panel, Regional and Language?

daj wrote:and Control Panel, Regional and Language?

http://www.meteosat.co.uk/Porthcawl/date.JPG

I think we've reached the end of this one - just wondering if the console has developed a fault?

Although Easyweather (NOT running at the same time) reads all the data off the console correctly and doesn't seem to have a problem - most strange!

Cheers

Steve can maybe pick up on this one when he is back -- tomorrow I think?

The strange dates in the diagnostics are a clue in this I think

Sorry I could not help more

daj wrote:Steve can maybe pick up on this one when he is back -- tomorrow I think?

The strange dates in the diagnostics are a clue in this I think

Sorry I could not help more

Not at all David - a sterling effort! - I hope we can find a solution soon and somehow perhaps marry the old data to the now very recent stuff. Looking in Easyweather, the console settings seem to be the same, but it is possible that I inadvertently touched something on the screen when updating the time - but for the life of me I cannot see what as the date and time format on the console appears to match all else.

Cheers

James

James
As a matter of interest, do you have any kids that might have thought it fun to change settings to see the effect??

What is your weather station date reading format (station not computer )??

As far as I know, Cumulus only uses the date and time from the PC and ignores the device

Super-T wrote:James
As a matter of interest, do you have any kids that might have thought it fun to change settings to see the effect??

No - just me and SWMBO !

It's definitely nothing to do with the station, it's something to do with the date settings on your PC. Windows is telling Cumulus that your short data format is m/d/yyyy - which in itself is a bit odd, as you might expect mm/dd/yyyy:

5/1/2010 17:58:10.948 : Short date format: 5/1/2010
5/1/2010 17:58:10.948 : Long date format: Saturday, May 01, 2010

I am also running Windows 7 x64, and this is what my date settings screen looks like:
I think that's the same as yours (as far as the date is concerned). Cumulus reads the date formats as:

02/05/10 12:48:39.822 : Short date format: 02/05/10
02/05/10 12:48:39.822 : Long date format: 02 May 2010

Note that both the short and long date formats are consistent with the control panel settings.

I don't understand how your date settings are being read incorrectly. How are you starting Cumulus? From the shortcut that it creates itself? You're running it in the same user as the one you used to take that screen shot of the date settings? You're not running it as Administrator, for example?

steve wrote:It's definitely nothing to do with the station, it's something to do with the date settings on your PC. Windows is telling Cumulus that your short data format is m/d/yyyy - which in itself is a bit odd, as you might expect mm/dd/yyyy:

5/1/2010 17:58:10.948 : Short date format: 5/1/2010
5/1/2010 17:58:10.948 : Long date format: Saturday, May 01, 2010

I am also running Windows 7 x64, and this is what my date settings screen looks like:

dateformat.png

I think that's the same as yours (as far as the date is concerned). Cumulus reads the date formats as:

02/05/10 12:48:39.822 : Short date format: 02/05/10
02/05/10 12:48:39.822 : Long date format: 02 May 2010

Note that both the short and long date formats are consistent with the control panel settings.

I don't understand how your date settings are being read incorrectly. How are you starting Cumulus? From the shortcut that it creates itself? You're running it in the same user as the one you used to take that screen shot of the date settings? You're not running it as Administrator, for example?

Hi Steve, and thanks for coming alongside.

What I did today was to start the program with the console connected (via the shortcut Cumulus creates) - which brought up the downloading data box - but then I just left it. I then discovered that after 20 minutes it had downloaded what data had accumulated since I had cleared the memory - which wasn't that much compared to what had been.
What I also saw in the last few minutes were all the figures and dials on the display quickly assuming a new amount, then moving onto the next, so it must be engaging with the prog, but at a very slow rate for some reason. Easyweather - when I tried that separately downloaded it in seconds and could display the graphs easily etc.

The only reason for all of this hair-tearing was because until now - if I needed to disconnect/reconnect the consols - as I have been experimenting with positions - the downloading took a few seconds, with the progress bar zipping by.

So currently it appears OK, but why it is now taking for ever to download data - I am not clear about.

Any thoughts?

Cheers

James

It's getting confused because of the two date formats and trying to download all of the data since January each time (or February, today, as it's now the second of the month).

steve wrote:It's getting confused because of the two date formats and trying to download all of the data since January each time (or February, today, as it's now the second of the month).

I think I must have completely messed up the console date - accidentally touching something.
Is there a way ro remedy this - will it heal itself at some point I wonder?

Cheers
James

Meteosat wrote:I think I must have completely messed up the console date - accidentally touching something.
Is there a way ro remedy this - will it heal itself at some point I wonder?

It's nothing to do with the console. Cumulus requires your system date format to stay the same, and somehow it appears to have changed. I really can't explain it, as your screen shot appears to show that your date format is dd/mm/yy, but when Cumulus reads it at start up, it gets m/d/yyyy. It's then comparing the date in month/day order with the stored date in day/month order and getting a mismatch. It thinks that the last time you ran Cumulus was several months ago, so it tries to download a huge amount of data from the station.

It's clearly not simply a bug in the system routines used by Cumulus on Win7 x64, as that's what I'm using and it works fine. I can only think that there is something unusual about the way that you run Cumulus (or perhaps in the way that you used to run Cumulus?). I'll see if I can come up with a test application for you to run to see if we can work out what's going on.

steve wrote:

Meteosat wrote: It's clearly not simply a bug in the system routines used by Cumulus on Win7 x64, as that's what I'm using and it works fine. I can only think that there is something unusual about the way that you run Cumulus (or perhaps in the way that you used to run Cumulus?). I'll see if I can come up with a test application for you to run to see if we can work out what's going on.

Thanks Steve,

Just to add - the PC runs continuously as it is processing satellite data. Until the problem occurred it had never been switched off - it is a new build, has no anti virus s/w as it never downloads, and any prog changes are passed over the LAN from a Kaspersky protected machine.

I also use perl scripts and MRTG.

An example is here:
http://www.meteosat.co.uk/mrtg/skystar_ ... e_snr.html

Note how it picks up the date etc.

Cumulus runs as suggested simply from the C: root drive and I have not altered the installation. I have tried renaming the directory and trying a fresh install to see if that helped, but it didn't.

Here is a couple of lines from the dayfile.txt

[General]
Date=5/3/2010
Timestamp=5/3/2010 9:00:00 AM
[Wind]
Speed=2.23693633079529

etc. (which is clearly the wrong way around.)

The file itself is marked 03/05/2010

The console reads 3.5.10 which is I believe correct.

I don't mind clearing the console memory again if it would help to sort things out once and for all.

Cheers

James

[EDIT] I have tried a small prog I used to use occasionally on an older XP machine to get UTC date time strings and on the Windows 7 64bit machine the results is the same format. Here is a pic:

http://www.meteosat.co.uk/Porthcawl/dateutc.JPG

Thanks to all inputting on this saga!

Cheers

James

Meteosat wrote:I don't mind clearing the console memory again if it would help to sort things out once and for all.

Excuse me for jumping in but, as I understand it, no console data can affect the date in Cumulus - so deleting console data won't help.

"I don't have the answer but I admire your problem". It's a strange one!