Hi Rune,

Ok, think I have spotted the issue. I mentioned that earlier versions of 
the gauge-data.txt template had a few issues; one of these issues was an 
error in the humidex formula that resulted in the same error as you have 
experienced when dewpoint approached 0 (C or F depending on your units in 
use). As dewpoint moved away from 0 the template would starting being 
produced again without error. I note from looking at your site your 
dewpoint has been around 0 at times today. I did submit a patch to the 
owner of the SteelSeries-Weather-Gauges GitHub repo sometime ago, which he 
did accept, but looking at the repo again just now it looks as if he 
reverted the change shortly after. I will take that up with him again.

In the meantime if you are up to editing gauge-data.txt.tmpl try replacing 
the following lines in gauge-data.txt.tmpl (circa lines 57-59):

#set $dp = $current.dewpoint.raw
#if $oT_C is not None and $dp is not None
#set $e = 6.11 * math.exp(5417.7530 * ((1/273.16) - 1/$dp))

with these:

#set $dp_K = $current.dewpoint.degree_C.raw + 273.15
#set $e = 6.11 * math.exp(5417.7530 * ((1/273.16) - 1/$dp_K))

and delete lines 69-71 inclusive:

#else 
#set $humidex = 'N/A' 
#end if

No need to restart, on the next report cycle the template should now 
generate without error, particularly when dewpoint is near 0. Let us know 
if that fixes it.

Gary



On Monday, 3 October 2016 21:38:02 UTC+10, Rune Lægreid wrote:
>
> Hi,
> I am using the gauge-data.txt.tmpl from the SteelSeries github site. To be 
> 100% certain I re- downloaded the file, but the issue is still there.
> It does not seem to create any problems, but it is kind of annoying seeing 
> it in the log.
>
> Rune.
>
> fredag 30. september 2016 15.47.17 UTC+2 skrev gjr80 følgende:
>>
>> Rune,
>>
>> The log extract below seems to indicate there is an issue with the 
>> gauge-data.txt.tmpl template which is part of the ss skin provided with 
>> the SteelSeries gauges v2.5.0 or later. gauge-data.txt (generated from 
>> gauge-data.txt.tmpl) 
>> was developed as the weewx standard data file used to feed the 
>> SteelSeries gauges when SteelSeries v.2.5.0 was reelased with native weewx 
>> support. Previously the only way to use the SteelSeries gauges with weewx 
>> was to mimic one of the other data files that the SteelSeries gauges 
>> accepts. This is what weewx-wd does, it produces the WeatherDisplay file 
>> customclientraw.txt via the SteelSeries skin. The upshot of it all is 
>> that wdSearchX3.py is only used by the weewx-wd skins, in this case the 
>> error is coming from a non-weewx-wd skin so the issue is not to do with 
>> wdSearchX3.py.
>>
>> Are you using the current gauge-data.txt.tmpl from the SteelSeries 
>> github site? I ask as the early versions had a few issues that under 
>> certain circumstances caused an error. I thought the current version had 
>> picked up all those issues, but maybe not. Once you confirm you are in fact 
>> using the lastest gauge-data.txt.tmpl I will have another look at the 
>> template and see if I can pickup where the issue is. If it turns out you ar 
>> enot using the lastest version I strongly suggest you update to the latest 
>> version. These sorts of errors from Cheetah are notoriously hard to track 
>> down, unlike the previous error which gave a line number in wdSerarchX3.py, 
>> this error gives a line number in a temporay file that we cannot access.
>>
>> Gary
>>
>> On Friday, 30 September 2016 18:01:43 UTC+10, Rune Lægreid wrote:
>>>
>>> Gary,
>>>
>>> That fixed this issue, but the logs indicates another issue with the new 
>>> wdSearchX3.py (I am, by the way one of the few with complete months gap in 
>>> my data):
>>>
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: cheetahgenerator: **** Reason: 
>>> math range error
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** Traceback (most recent call 
>>> last):
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** File 
>>> "/home/weewx/bin/weewx/cheetahgenerator.py", line 300, in generate
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** print >> _file, text
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** File 
>>> "/usr/lib/python2.7/dist-packages/Cheetah/Template.py", line 1005, in 
>>> __str__
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** rc = getattr(self, 
>>> mainMethName)()
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** File 
>>> "_home_weewx_skins_ss_gauge_data_txt_tmpl.py", line 142, in respond
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: **** OverflowError: math range 
>>> error
>>> Sep 30 09:53:05 WeatherPi weewx[27820]: cheetahgenerator: Generated 1 
>>> files for report SteelSeries in 6.27 seconds
>>>
>>> mandag 26. september 2016 01.28.56 UTC+2 skrev gjr80 følgende:
>>>>
>>>> My apologies but I discovered another bug in the code. It will only 
>>>> affect you if you have a complete months gap in your data so its unlikey 
>>>> to 
>>>> impact many folks. The fix is the same as the link I posted earlier but 
>>>> you 
>>>> need to get wdSearchX3.py as follows:
>>>>
>>>> wget https://
>>>> bitbucket.org/ozgreg/weewx-wd/raw/6344acab608408b5826d52b1d4e84667220c09de/bin/user/wdSearchX3.py
>>>>
>>>> Gary
>>>>
>>>> On Monday, 26 September 2016 08:50:46 UTC+10, gjr80 wrote:
>>>>>
>>>>> Hi,
>>>>>
>>>>> You are correct, the issue has come up a couple of times before. The 
>>>>> code on bitbucket has been fixed for sometime but I have not packaged up 
>>>>> the patched file. My bad, will make a point of doing that today. The 
>>>>> easiest fix for you is to follow the instructions here 
>>>>> <https://groups.google.com/d/msg/weewx-user/j5qfHK3P_ME/shRRmJDzAQAJ>.
>>>>>
>>>>> Gary
>>>>>
>>>>

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to