After spending some time fixing cwxn to address the issues I found, I took 
a brief look at weewx-aprx.  I forget exactly why I decided not to go that 
route, I think it pertained to the way it was generating the packet which 
didn't entirely mesh with how I wanted it for direwolf.  So then I decided 
to leverage the CWOP support since it's basically the same format.  I 
enabled the CWOP support (so that it's sending directly to aprs-is) and I 
made a couple minor edits in restx.py to output the relevant bits of the 
packet to a wxnow.txt which direwolf can use.  Unfortunately I just did 
them as local edits - I didn't pull the weewx branch and make the changes 
via git.  And I basically just hard-coded the changes in.  What I should've 
done was add an option in the weewx.conf to write it out and then maybe 
submit a pull request to get that included in the weewx sources.  

Depending on your comfort level dealing with source code, if you want, I 
can send you my restx.py with my changes, but you'll want to compare with 
your restx.py in case it's not based on the same version.  

-Dave

On Monday, November 23, 2020 at 5:28:26 PM UTC-8 [email protected] wrote:

> Hello - 
>
> Longtime Weewx user trying to get his station on APRS (RF).
>
> I'm using Direwolf and, ideally, would just like to get a wxnow.txt file 
> that it can pick up.  However, it appears that the original author of the 
> cxwn extension may have pulled it down.
>
> I see a reference to another plugin -  weewx-aprx  - but I'm unable to 
> locate that as well.
>
> Any assistance would be great!  Thanks!
>
>  
>
> On Thursday, September 17, 2020 at 12:09:12 PM UTC-4 [email protected] 
> wrote:
>
>> Gary,
>>
>> Thanks!  I just took a look at the weewx-aprx extension.  FWIW, in 
>> addition to fixing the humidity, I also added solar radiation 
>> (luminosity).  One other thing to note which I'm likely going to try and 
>> fix shortly, and you might consider fixing in weewx-aprx, is that the 
>> nullproof probably shouldn't turn those values into 0.  To indicate "no 
>> value", I believe periods should be used - otherwise it's interpreted as a 
>> real value of 0. 
>>
>> Here's the code I added for radiation:
>>
>> # DWC adding solar radiation. If it's below 1000, use 'L' 
>> # if it's above 1000, use 'l' and provide the value above 1000 
>> if data['radiation'] < 1000: 
>>     fields.append("L%03d" % int(data['radiation'])) 
>> elif data['radiation'] < 2000: 
>>     fields.apped("l%03d" % (int(data['radiation'])-1000)) 
>> else: 
>>     loginf("radiation looks bad: %d" % int(data['radiation']))
>>
>> -Dave
>>
>> On Thursday, September 17, 2020 at 3:16:12 AM UTC-7 gjr80 wrote:
>>
>>> Thanks, subsequent to the above thread and in concert with Bill I created 
>>> the weewx-aprx extension <https://github.com/gjr80/weewx-aprx> which 
>>> does include the correct humidity behaviour.
>>>
>>> Gary
>>>
>>> On Thursday, 17 September 2020 at 11:16:08 UTC+10 [email protected] 
>>> wrote:
>>>
>>>> Sorry for bringing an old thread up here.  I just posted this to the 
>>>> related wxforum post:  https://www.wxforum.net/index.php?topic=38877
>>>>
>>>> I'll copy-paste my post here:
>>>>
>>>> Following this thread and some threads and references elsewhere, I just 
>>>> got my WX station working with direwolf and weewx running on an rpi, and I 
>>>> wanted to point out that there's a small issue with the cwxn.py 
>>>> implementation, at least for APRS use.   
>>>>
>>>> At first I thought everything was going great - I was beaconing out the 
>>>> wxnow.txt info and looking at aprs.fi everything looked good.  But 
>>>> then I looked at the info in findu and noticed the dew point seemed to be 
>>>> messed up.  Looking at it more closely, that was because the humidity was 
>>>> wrong.  Instead of, for example, 92%, it was showing up as 9%.  So I took 
>>>> a 
>>>> closer look at the format of the data.  The humidity is being formatted as 
>>>> a 3 digit number by cwxn.  But the APRS spec and CWOP spec indicate it 
>>>> should be 2 digits, and 00 should be interpreted as 100%.  Apparently 
>>>> aprs.fi is being tolerant of the 3-digit number, but findu is not. 
>>>>
>>>> I made a small change to cwxn.py to convert 100 to 0 and to format as 
>>>> %02d instead of %03d.  All looks good in both aprs.fi and findu now. 
>>>>
>>>> BTW, in case it's relevant, I'm using the cwxn.py that was posted 
>>>> earlier in this thread by gjr80.  
>>>>
>>>> My station:
>>>> https://tempestwx.com/station/28304
>>>> https://www.wunderground.com/dashboard/pws/KCAPACIF205
>>>> http://www.findu.com/cgi-bin/wxpage.cgi?call=AE6DC-5&last=240
>>>> https://aprs.fi/info/a/AE6DC-5
>>>>
>>>> -Dave AE6DC
>>>>
>>>> On Wednesday, March 4, 2020 at 6:09:41 PM UTC-8 gjr80 wrote:
>>>>
>>>>> That is good news, the cwxn code is fairly basic and should have 
>>>>> worked on your earlier system.
>>>>>
>>>>> Gary
>>>>>
>>>>> On Tuesday, 3 March 2020 11:45:10 UTC+10, Bill Arthur wrote:
>>>>>>
>>>>>> Gary,
>>>>>>
>>>>>> Thanks again. Everything is working perfectly. 
>>>>>>
>>>>>> BTW, I installed Weewx on a new RasPi.  "binding  = archive" works 
>>>>>> perfectly
>>>>>>
>>>>>> Bill
>>>>>>
>>>>>> On Sunday, March 1, 2020 at 9:36:30 PM UTC-6, Bill Arthur wrote:
>>>>>>>
>>>>>>> Gary,
>>>>>>>
>>>>>>> I watched the results after the last rain. The midnight column went 
>>>>>>> to 0.0 at midnight and the 24hr went down after 24 hours. Everything as 
>>>>>>> expected.
>>>>>>>
>>>>>>> It has now just started raining and all three columns are at 0.01.
>>>>>>>
>>>>>>> I believe we can say that the problem has been resolved. Thank you 
>>>>>>> for your quick and precise solution.
>>>>>>>
>>>>>>> Bill
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Saturday, February 29, 2020 at 12:21:26 AM UTC-6, Bill Arthur 
>>>>>>> wrote:
>>>>>>>>
>>>>>>>> Going through the log it looks like I may have restarted the box 
>>>>>>>> when I installed an app. I didn't think I did it that early in the 
>>>>>>>> day, but 
>>>>>>>> I may have. I'll take the blame. 
>>>>>>>>
>>>>>>>> Since then it's working flawlessly, the midnight total went away at 
>>>>>>>> 00:00. 
>>>>>>>>
>>>>>>>> I'll know more Monday when it rains. 
>>>>>>>>
>>>>>>>> Bill 
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Friday, February 28, 2020 at 8:55:59 PM UTC-6, gjr80 wrote:
>>>>>>>>>
>>>>>>>>> Don’t get too hung up on wxnow.txt, it merely pulls data from the 
>>>>>>>>> database/the current archive record and presents it. If the WeeWX 
>>>>>>>>> generated 
>>>>>>>>> plot you posted is correct then there is a discrepancy between what 
>>>>>>>>> WeeWX 
>>>>>>>>> is receiving and what your meteobridge is receiving. This is best 
>>>>>>>>> tracked 
>>>>>>>>> down by looking at the underlying data.  Once we know the correct 
>>>>>>>>> data is 
>>>>>>>>> being received/archived by WeeWX we can follow that data through 
>>>>>>>>> WeeWX. 
>>>>>>>>>
>>>>>>>>> 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].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/weewx-user/53d48c75-bc0f-4d34-9e2d-9406eb95980en%40googlegroups.com.

Reply via email to