Closed meirion closed 3 years ago
You probably had wifi dropouts so your live graph didn't get the event notifications.. if you refresh, i think it should display
See #199 - it's the same issue. Refreshing probably won't help in this case since the updates never make it to the graph. Refreshing does help sometimes if the browser isn't paying attention, say if your computer goes to sleep or something.
The devices only send the "change step event" once and never listen to a server response... So yeah if we miss processing these (see https://github.com/chiefwigms/picobrew_pico/issues/199 for some issues... I'm working on an upgrade for that issue stay tuned) you will be missing them.
We could just mark the step upon first data point we receive, but that would be lying to you when the step started so that wouldn't be cool and thus we don't do that.
Been busy here with work this week and was fixing some tilt functionality last week after merging with Charles. So long story short haven't gotten around to adding the upgrade script change for https://github.com/chiefwigms/picobrew_pico/issues/199, but the config changes have been merged into the repo and should be fairly easy now with the upgrade after git pull script we have added.
Ah I understand, thanks for the explanation.
Are there steps for me to improve the WiFi? I have the pi next to the Pico currently.
The suggestion of dropping a lien the first time you see the data for a new step could be interesting- maybe display in s different style to indicate some data was missing... Dotted or greyed perhaps.
That would be useful as users would then have an indication that their log data was not complete, and that their WiFi is missing some.
For example I did not know that was occurring until I asked on here.
Thanks
Yeah could think about showing "data is missing or incomplete" or something when we detect that.
Actually the wifi isn't what is resulting in a number of missing events, but rather the nginx config.
Give me 2 days and I'll get an automatic way to get this known issue resolved.
In my case, updating the nginx configuration helped, but didn't solve the problem - only switching to two WiFi interfaces fixed it. I'm curious if it will be helpful to people with more marginal WiFi configurations.
@zuiko42 I've had 2 others show massive improvements with just the nginx configuration. Depending on the rssi from the device to the Pi I'm sure there could still be drop outs, but we have nearly definitely shown that the Pico devices don't actually keep the connection open listening for a response all the time (thus the client 499 error we saw) seems to be implemented in a fire and forget manner.
Really confused by the marker lines on the Graph - some from the recipe show, but some are missing. On a previous brew I saw more marker lines (different recipe)