On Thursday, January 21, 2021 at 6:53:11 PM UTC+1 [email protected] wrote:

> On Wednesday, January 20, 2021 at 11:10:35 PM UTC-8 Jan-Jaap van der Geer 
> wrote:
>
>> > * if you're asking if there's some kind of hybrid UDP 'and/or' REST 
>> weewx driver...
>>
> I'm not asking that, but the latter is what I'm trying to create. Well, 
>> it's working though very much a WIP. However I have wondered as well if I 
>> should drop the UDP altogether and end up with a REST-only-driver. I 
>> haven't really decided on that yet. Maybe I'll make it configurable in an 
>> all-in-one driver.
>>
>>
> Nobody's going to turn down having more drivers available :-)
>

No? :) 

I guess I could see a few options:
>
>    1. a REST-only driver
>    2. a driver that is configurable UDP 'or' REST
>    3. a driver that is smart enough to use REST to fill in UDP gaps
>    4. or maybe even something like wunderfixer as a standalone utility to 
>    fill in weewx gaps by comparing the db with the WF server data
>
> Never looked into wunderfixer but heard the name, sounds interesting. I 
suppose it isn't something general that works with the current drivers?

Anyway, the goal is nr 3 although it will only look at the gaps since the 
last measurement, so not really gaps.
 

> Power loss risk is easier.  Buy a UPS.  Put your Hub on it.
>

My hub is on an old 10.000 mAh battery. No idea how long of a power loss it 
would survive (but then that's not really a big issue usually here anyway). 
Hm, did I introduce a fire risc?

Old posts on the WF Forums from their engineering said that the old Air+Sky 
> sensors could store 90-120 minutes if they can't reach their Hub.  The Hub 
> stores 8 days of data if it can't reach WF servers.   So if you're 
> concerned re: short outages then I'd UPS the Hub and ideally the weewx box 
> if you could.    The new Tempest stations are similar enough those numbers 
> are probably reasonable for the current gear.
>

Ah, that's a lot more than I thought. I thought I'd seen something like 1h 
for the hub, and I was assuming nothing for the sensor. Sounds good! 

Personally I kinda like the idea of (1) and (4) at least initially, and 
> maybe (2) as a stretch goal.
>

Well, aiming for (3) for the time being... :-P

Cheers,
Jan-Jaap

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-development" 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-development/63b27c61-0d5b-41a2-8fcc-71a5f95d42c6n%40googlegroups.com.

Reply via email to