That seems to have fixed the problem, thanks Gary.
On Wednesday, 25 October 2017 09:57:35 UTC-4, Robert Mantel wrote: > > Ok, I will try this, and yes I should update the extension....thanks for > the heads up. > > On Wednesday, 25 October 2017 09:24:17 UTC-4, gjr80 wrote: >> >> The error below is due to a bug in the trend calculation code in the >> realtime gauge-data extension. In this case the historical observation >> being used in the trend calculation is None and that is throwing the error. >> I will raise an issue and fix it in the next realtime gauge-data extension >> release. In the meantime if you are up to it you can make the following >> change to bin/user/rtgd.py line 1956 fix the problem: >> >> if then_record is None: >> return None >> else: >> if obs_type not in then_record or then_record[obs_type] is None: >> return None >> else: >> then_vt = weewx.units.as_value_tuple(then_record, obs_type) >> now = convert(now_vt, group).value >> then = convert(then_vt, group).value >> return now - then >> >> You will need to restart weeWX after saving the change. >> >> It looks as if you are using v0.2.10 of the extension, the current >> release is v0.3.0 which fixes a number of issues (not the one above) and >> adds a number of new features. >> >> Gary >> >> >> On Wednesday, 25 October 2017 11:55:40 UTC+10, Robert Mantel wrote: >>> >>> Hey Fredric, I implemented your version of the wxMesh driver, but >>> sometimes I'm getting this error and I'm not sure if it's the realtime >>> gauge data issue or the wxMesh issue but sometimes I'll get multiple cycles >>> of this: >>> >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** Traceback >>> (most recent call last): >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** File >>> "/usr/share/weewx/user/rtgd.py", line 875, in process_packet >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** data = >>> self.calculate(cached_packet) >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** File >>> "/usr/share/weewx/user/rtgd.py", line 1030, in calculate >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** >>> self.db_manager, ts - 3600, 300) >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** File >>> "/usr/share/weewx/user/rtgd.py", line 1962, in calc_trend >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** return now >>> - then >>> Oct 24 21:48:22 raspberrypi weewx[1298]: rtgdthread: **** TypeError: >>> unsupported operand type(s) for -: 'float' and 'NoneType' >>> >>> Now I'm getting updates from my Moteino powered weather station about >>> every 3-4 seconds and have wxmesh setup to take loop packets as quickly as >>> it gets them, I'm wondering if it's getting out of sync somehow? Maybe >>> someone else has seen this? Over all this works great. I love having the >>> information available in mqtt which makes it usable not just for weewx but >>> for anything else where you'd like to use that data. >>> >>> On Monday, 7 August 2017 09:56:08 UTC-4, Wysiwyg wrote: >>>> >>>> Hello, >>>> >>>> The initial driver from Bill is working (as far as I know). >>>> It is subscribing to a single mqtt topic which should contain a long >>>> string looking like this: >>>> TIME:0,AMBT: 7.23,BARP:1001.97,RHUM:70.92,HUMT: 4.44,IRRA:12,BATV:974, >>>> PHOV:673,SYST:33.20,WIND: 0.0,WDIR: 0.0 >>>> >>>> In the conf file, it is set the dictionnary to convert labels like >>>> "HUMT" to weewx database schema 'outHumidity'. >>>> >>>> >>>> >>>> My own modification if based on Bill's job, but I modified a bit to use >>>> separated mqtt topics for each sensor. >>>> for example I have the following mqtt topics: "weewx/outTempature" , >>>> "weewx/barometer" etc... >>>> the sensors publish the measurement value a string in the topics. for >>>> example "23.42" for temperature (when it's 23.42°C). >>>> weewx is subscribing to all topics below "weewx/" (but you can choose >>>> something else as root topic). >>>> >>>> I have done this modification because my sensors are at different >>>> locations and different rates. >>>> for example, temperature is published every few minutes, rain is >>>> published only when it's raining. >>>> >>>> to make it simple, as I am a python noob, I use as right part of the >>>> mqtt topic exactly weewx database schema : >>>> for example, in the topic " weewx/outTemperature" the "outTemperature" >>>> is the same label as weewx use. >>>> >>>> in case you publish some topics not available in the schema, weewx will >>>> just not consider them. >>>> For example, I have published some topics like "www/Vsolar" which is >>>> the voltage of the solar panel supplying the sensor. >>>> This value Vsolar is rejected by weewx unless you add it in the schema. >>>> >>>> This is working pretty well for some time now on my setup (barometer, >>>> inTemperature, outTemperature, outHumidity, rain & UV are working well.) >>>> >>>> >>>> >>>> >>>> I hope that helps ? >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> Le lundi 7 août 2017 14:21:57 UTC+2, Neil Trimboy a écrit : >>>>> >>>>> I am wanting to get some remote sensors data into weewx and was >>>>> considering mqtt. >>>>> It seems there are currently at least 2 threads of development on >>>>> drivers/services to load data into weewx ? >>>>> >>>>> Could someone summarize their differences and statuses please?. Are >>>>> they forks and work fundamentally differently, or is one a more mature >>>>> version ? >>>>> >>>>> thanks >>>>> >>>>> On Friday, 7 July 2017 07:10:03 UTC+12, Wysiwyg wrote: >>>>>> >>>>>> Hi There! >>>>>> >>>>>> I just wanted to share a bit the news on this project. >>>>>> >>>>>> The mqtt driver works since a few monthes now with pressure, >>>>>> barometer & intemp like a charm ! >>>>>> Since a few days, I introduced outTemp, outHumidity, rain & UV. it >>>>>> works perfectly ! >>>>>> >>>>>> Best regards, >>>>>> Frederic >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Le jeudi 6 avril 2017 22:58:22 UTC+2, Wysiwyg a écrit : >>>>>>> >>>>>>> Thx for feedback Andy !! >>>>>>> Nice to hear it works fine ! >>>>>>> >>>>>>> the MQTT driver works fine also on my side, with barometer & inTemp >>>>>>> sensors. >>>>>>> Still slowly but not stopping developping the other sensors >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>
