Weather34 Aurora Version(english supported only)

[jasonmfarrow]

... all done but note Firefox,Safari,Chrome are beginning to render some code differently so beware there may be some minor position difference in these browsers...

Brian, absolutely. This latest update has the close button aligning nicely to the top right on Safari, but has a larger right gap on Chrome. This must be so frustrating as a designer and difficult, if not impossible, to fix. You'll please some and upset others with the fix!

[weatherist34]

bart i have to go but your airlink is definitely NOT being read the nanosd my gut feeling running two meteobridge devices simultaneously into one sensor may be the issue i cant confirm but looks problematic

[Bart]

Thanks Brian. Will look into that. The second site I installed quick and dirty fast. So that delivered the problem. But the NanoSD one ran for a while. No changes where made there. At least temporary I will remove the air link from the RPI site and see what will happen on the first (NanoSD) one.
Thanks for you time. Have a great time!
Bart

[Bart]

Brian, I've looked at my setups here. I solved the AQ issue. My problem which arose on the NanoSD was twofold.
Too many commands for the NanoSD and the Nano had problems with my Network. In a mesh on the 2.4 GHz. That went well for a long time, but I think an update somewhere caused problems with the connections.
So I set up that NanoSd on my guest network. Although part of the whole (of the same LAN) but still a different WiFi network. 2.4GHz only.
And then it turns out that the NanoSD cannot reach the AirlLink, which is in the WiFi of my main network. A pretty silly problem, the Nano is the only device that can't cross-communicate in the network. All my other WiFi connected devices (and there are a lot of them) except for one other, can do that, just my vacuum cleaner doesn't......
Bad of Meteobridge in fact, but not insurmountable, once you realize it.

So it is not so much that the information from the Airlink cannot be shared twice or more, but a problem of connectivity.

The Meteobridge RPI is wired into the network and can of course simply reach the Airlink, which is also in the main network.

Just thought I'd report this to you. For future reference. You never know...

[vortexya]

Gday Brian, Just a heads up that the latest update looks fantastic on Pale Moon browser, and on the 65inch Plasma I use for a monitor
Lockdowns here too mate,the land of dropbears is now into winter, at least the rain gauge is getting a good workout I guess. All the best, Gary

[jmontamat]

Hi Brian, I had to adjust a bit

.lity-close {right: 2.15em}

So that it works well in w10 and androd browsers.

as you predicted, now it looks better.

[jasonmfarrow]

Brian, Two things to think about.
1) I prefer the "close" button in the absolute position two right of screen. It won't keep still with different browsers in the relative position of the open window.
2) The day-rain chart scale appears to be based on the 24-hr_total not the current day_total.

Capture13.PNG (18.37 KiB) Viewed 3568 times

All the other day-charts' scales are proportional (+/-) to the current values. I saw a particularly bad example of this before where 0.2mm of rainfall that had fallen recently was not visible on the chart because the scale used the 24-hr total (47mm). [This may well disappear when you finish your re-work of the chart code....]
Hope you're enjoying the time away...

[jmontamat]

In the last update it is seen there:

[jasonmfarrow]

Nope. It's not the chart. It's the raw data. In weather34charts/2021/5Jul2021.csv the first few lines are:

Code: Select all

0,00:00,13.9,1007.3,11.0,0.0,4.5,1.8,0.0,12.8,0.0,234.0,22.7,65,5 Jul,13.5,--,93,4.2
0,00:10,13.9,1007.2,0.0,0.0,4.0,1.5,0.0,12.8,0.0,234.0,22.7,65,5 Jul,13.5,--,93,3.3
0,00:20,13.9,1007.2,0.0,0.0,3.6,1.8,0.0,12.8,0.0,188.0,22.7,65,5 Jul,13.5,--,93,2.1
0,00:30,13.9,1007.1,0.0,0.0,4.9,2.0,0.0,12.8,0.0,240.0,22.7,65,5 Jul,13.2,--,93,3.8
0,00:40,13.8,1007.1,0.0,0.0,4.5,1.4,0.0,12.7,0.0,186.0,22.7,65,5 Jul,13.7,--,93,3.8
0,00:50,13.7,1007.3,0.0,0.0,3.6,1.8,0.0,12.6,0.0,202.0,22.6,65,5 Jul,13.2,--,93,2.5

The 5th comma delimited item is the day-total rainfall. However this was yesterday's day-total rainfall. The rainfall at midnight should be zero...but it's not. Changing that 11.0 to 0.0 results in this chart:

Screenshot 2021-07-05 at 5 July 2021 20.41.17 pm.png (30.3 KiB) Viewed 3542 times

So there are two possible fixes: 1) correct the chartstoday.php code to use the correct value at midnight (00:00) or 2) get the weather34charts/todayrainfallmodulechart2.php chart code to ignore midnight rainfall readings.

[jasonmfarrow]

Considering that the chartstoday.php script is scheduled by my Meteobridge NanoSD "every full 10 minutes" there is a case for saying that the bug might exist at the Meteobridge end. According to Brian's instructions chartstoday.php is called by the meteobridge "every full 10 minutes".

If "every full 10 minutes" adheres to our expectations then it should execute any script at precisely 23:50:00 and 00:00:00 and 00:10:00 etc. Which then, if true, at 00:00:00 the Day_Rain_Total should be zero. If the "every full 10 minutes" is not precisely 00:00:00 but might be 23:59:59.9 then I can understand why the Day_Rain_Total is not zero when calling chartstoday.php.

@Boris, can you comment on the precise timing of "Every_Full_10_Mins" in the Services->Events section on the Meteobridge?

[jmontamat]

I use TP-link and the first record is at 00:09 maybe it is a time lag between meteobridge and the server template?

[jasonmfarrow]

I use TP-link and the first record is at 00:09 maybe it is a time lag between meteobridge and the server template?

There is a difference between "Every 10 Minutes" and "Every Full 10 Minutes". You use the former; the timing of which is established from when the meteobridge was last rebooted. "Every Full 10 Minutes" is supposed to be 00,10,20,30,40 and 50 minutes past each hour.

[jmontamat]

You are right I will try Every Full 10 Minutes in my test page to yer what happens tomorrow .....

https://meteosantgenis.org/pws1/

[jasonmfarrow]

You are right I will try Every Full 10 Minutes in my test page to yer what happens tomorrow .....

https://meteosantgenis.org/pws1/

You have clearly implemented this "every full 10 minutes" and demonstrated the problem. You had 1.1mm of rain yesterday. This is reflected in your charts and it's making the mess of the current chart as your 8July2021.CSV has 00:00 rainfall at 1.1mm when it should be zero...

Capture14.PNG (40.93 KiB) Viewed 3453 times

This shows that for some reason the 00:00 chartstoday.php is using the previous day's Rain_Day_Total and not resetting to zero...

[weatherist34]

Considering that the chartstoday.php script is scheduled by my Meteobridge NanoSD "every full 10 minutes" there is a case for saying that the bug might exist at the Meteobridge end. According to Brian's instructions chartstoday.php is called by the meteobridge "every full 10 minutes".

If "every full 10 minutes" adheres to our expectations then it should execute any script at precisely 23:50:00 and 00:00:00 and 00:10:00 etc. Which then, if true, at 00:00:00 the Day_Rain_Total should be zero. If the "every full 10 minutes" is not precisely 00:00:00 but might be 23:59:59.9 then I can understand why the Day_Rain_Total is not zero when calling chartstoday.php.

@Boris, can you comment on the precise timing of "Every_Full_10_Mins" in the Services->Events section on the Meteobridge?

Hello sorry not been able to engage , the problems lies in meteobridge itself has it sends yesterday rain total at midnight as you can see in your csv file at 10 minutes after midnight its not visible . template simply only reads what it is in the file and the data is what is sent from meteobridge itself at that given time.

so ive added a simple ignore first line of csv data in the raintoday chart therefore in raintoday chart the data visible starts at 00:10 not 00:00 ,all we are doing is disguising the anomaly.

there are some other small changes in the css file .

i will push the update sometime later today , back in istanbul this morning until monday then im off again till end of month as its a national holiday week end of month.

so thats it for these charts using canvasjs im absolutely not doing anymore as i will just focus on new chart library when i can find the time , at present we are just catching up with all the things we couldn’t do for the last year or so i.e travel,see relatives etc ..

reference close icon position yes i agree its better to have a fixed position as it 99% guarantees universal appearance across devices,browsers etc so ive reverted to original fixed position.

kindly ask no more request,bugs based on using this canvasjs library it is what it is for now and sometime soon they will not be used and removed.as before canvasJs is powerful but complex newer charts makes everything simpler to maintain and design around and most importantly a very low overhead in resource use. note new chart library you will be able to enlarge charts when viewing in either extra info or chart page the larger charts will simply be viewable with a simple click in a extra popup so you will have normal and larger viewable options on the fly.

as above will push,upload update sometime later today.

ps ecowitt users im so sorry but there is 100% no planned support for now too much confusion and no clear cut methods for supporting
some of the additional sensors i.e specifically lightning ,simply no point until a common method has been defined .this is not my problem so back off trying to pinpoint weather34 template is the issue , i say again weather34 does not support ecowitt or alike at this point in time so there not template issues as ecowitt is not developed for.

have a good day.. brian ps torsten i have a solid solution for rain last tip again its a disguise due to meteobridge does not send a valid timestamp after 72 hours or 3 days if no rain .sorry about the football