Cumulus Support

I have updated to v.1.2 and all is well. I also now better understand the data provided by Cumulus or calculated versus the data provided from the Metar. I had in error thought that by setting '$useCUHeight = false' and '$onlyUseMetar = true' would get the Metar cloud height, but I now see that Metar does not provide the cloud height in their hourly string (it does show on their WU page and presume that comes from some highly technical equipment) and that the cloud height comes from either Cumulus or script calculated. I have the settings $getMetar, $useMetarAtNight and $onlyUseMetar all as false and the cloudbase image seems fine.

Again thanks,
Paul

Yes, as per the notes beside the settings, if $getMetar = false; then there is no metar functionality.

As you now have Metar turned off, I would appreciate feed back in due course on how reliable you find the Zambretti alternative.

I have updated to v 1.2 and it's working fine (also windsock is "moving" when wind speed changes). What I wander is why there is so big difference between wind speed (see attachment) shown by this script and the Ray's "ancient" script ( I have installed both just for comparison). Do they have different source of data (I have not compared "realtime.txt" and "cloudbaseCUT.txt").
Best Regards,
Piotr

Sorry, I can't answer that (no longer know how the 'other' script works and don't have the files anymore).
Just checked your current image with ?html=1 and graphic is same as data coming from Cumulus.
We have steady conditions at the moment and dashboard realtime wind speed is the same as cloudbase wind speed.

CloudbaseCU.php is updated by FTP at whatever interval that is at (10 mins for our site) - cloud conditions don't change very fast
realtime.txt is usually updated much more frequently (15 secs for our site) so maybe the wind speed was different at the time the different files were uploaded.

As you now have Metar turned off, I would appreciate feed back in due course on how reliable you find the Zambretti alternative.

So far it looks ok. I'll watch it and let you know.

Paul

BCJKiwi wrote: CloudbaseCU.php is updated by FTP at whatever interval that is at (10 mins for our site) - cloud conditions don't change very fast
realtime.txt is usually updated much more frequently (15 secs for our site) so maybe the wind speed was different at the time the different files were uploaded.

Thanks. I should have been expected such an answer (in regards to "ancient" cloudbase script) to my stupid question.
Just for testing purpose I set realtime update for "cloudbaseCUT.txt" file and noticed that difference. Of course, wind speed data is not important in this script

Best Regards,
Piotr

Well, am in the middle of quite a heavy thunderstorm but, apart from a thickening of the cloud on the image and condition of "Overcast" there doesn't appear to be anything else to indicate that. Should there be?

So far it looks ok. I'll watch it and let you know

Well now is a good time to let you know. It is still very sunny with cumulus clouds in the sky (check webcam image on http://www.komoakweather.com ) but shows "CUMULUS CONDITIONS Zambretti # : 15 Current Cloud conditions : drizzle,shower". Looks like I will have to experiment some more but currently not at home and will have little time before leaving on a 10-day cottage vacation tomorrow morning. I do notice that presently shows cloud height of 1072 m which is much higher than what I have seen it calculated over the past few days. Perhaps this influences the Zambretti forecast?

For what it's worth this is the Davis station's forecast "Forecast: Increasing clouds with little temperature change. Precipitation possible within 24 to 48 hours"

I may have to reset it to use Metar for now.

Paul

Glupack,
The script can only work with the data it receives. So when something like this happens, check with cloubaseCU.php?html=1 to verify the script is working properly.
If it's not, please post the image (It is saved as cb.jpeg in the script folder when you use ?html=1) and a listing of the text produced by ?html=1 so I can have a look.

Oh, I see you are using Zambretti rather than Metar - refer to other discussions on the difference. Functionality without Metar is not nearly as representative of actual conditions in my experience.

Thanks

Hi, yes I have checked .php?html=1 and the data is correct to what is produced within Cumulus so the script is working fine and wanted to give you feedback on the Zambretti alternative. I may experiment with the 'switch ($cond) re Zambretti forecast number as you had indicated in an earlier suggestion, and that is a whole new territory for me but very helpful in learning/understanding the whole process. For now back to the Metar source.

Thank you for your follow up and continued work on this,

Paul

The discussion re the update time has prompted me to again let people know how I have things set up regarding meta refresh.
As mentioned above, our FTP updates are at 10 min intervals = 600 seconds.
Cumulus synchronises the FTP intervals to the hour (good to have the FTP interval an equal divisor of 60 minutes).
So some time ago I looked for a way to get meta refresh to synch to the updates.

This is what I came up with;
In top.php (Saratoga template), the first line under <head> is,
<meta http-equiv="refresh" content="<?php echo $refresh; ?>" />
so instead of a number of seconds we have <?php echo $refresh; ?>

To get a number into $refresh we calculate the time to the next refresh so that if you open the page at some random time, or manually refresh it at some random time, it will still 'meta refresh' just after the data is uploaded.

In CUtags.php immediately before "//end of generation script" $refresh is give a value by this code;
// synchronise page refresh to tags upload time
if(isset($SITE['WXtagsPeriod'])) {$period = $SITE['WXtagsPeriod']+15; // $SITE['WXtagsPeriod'] = secs between "WXtags" file uploads. +15 to ensure data has uploaded.
$age = time() - filemtime($SITE['WXtags']); // now - time CUtags.php was modified
if($age < $period) {$refresh = $period - $age;} else {$refresh = $period/2;}
}

"$SITE['WXtagsPeriod'] = xxx ;" is set in Settings-Weather.php in the section "# required settings for Cumulus software"

I'm sure there are different ways this could be done but I have not seen it done elsewhere that I could find.

Caveats;
I know many like to knock Microsoft, and specially IE, but this works well in IE8, 9 & 10
It does not work in Chrome as Chrome does funny things with meta refresh (it uses a different layout engine from IE but I don't know if this is the cause).
I have not tested this in other browsers.

Hi
A little bit o.t. but when we get that Davis VP2 console script "made for Cumulus"?

patience, patience!

It seems to me that there are only two letters of wind direction shown on the picture (see attachment). I noticed that couple of times with different wind directions. Sorry for still using "ancient" script for comparison. Both scripts are working in realtime mode.
Best Regards,
Piotr

What happens if you run it in English?
I don't recognise your translation and changing to English does not seem to work?

?html=1 at the time of this post shows NW as the wind direction.

There are diagnostics in the script to help you with resolving such questions

Have tested the script with text the same length as your sample and don't see what you indicate in your post. The wind dir should always be in the same position as there is a calculation to position it taking into account the length of the text.
Can you pm me your current cloudbaseCU.php file so I can see why your script is not behaving the same as ver CU1.2 as published.