OK, so I have reverted to the unsecured websocket and have added the extra debug output. Can you patch again using wfpiconsole patch and then get the console back up and running. Now you should see a message printed every time a Sky message is received, along with the time since the last message in brackets. I am interested in know what happens when the console freezes.
Before you re-start the console, can you also run python3 -m pip list --format columns and copy/paste the output. That will tell me what versions on the different Python modules you are using. I can compare that with my output to see if you are using a different (and maybe buggy) module.
Also don’t worry about the extra output once you have pressed ctrl-c. That is a standard error indicating that you have used a keyboard interrupt to stop the console/copy the text. You might be able to right click on the text in VNC as well to copy instead.
#Peter first rain / a few lightnings since we upped to this version and definitively something fishy
As you know I’m using the Tempest. And is indeed like frozen, no more wind, no pressure update etc …
You too! There clearly is same nasty bug that needs squashing. I was having a think this evening and I might have an idea as to what it is. Let me make a new patch tomorrow and then I’ll get you to update to that.
@padieter Glad to see the patch is working. I am really interested in knowing whether the Sky messages continue to arrive after the console stops updating. This will really help clear up whether it is due to a loss of connection with the servers, or something internal to the console code. I am hoping for the latter right now, as I might have realised what the issue is.
@eric and @padieter, I have just pushed another patch to Github (v3.1.4), which I am really hoping will fix the freezing issues. Can you both run wfpiconsole patch and manually start your consoles with wfpiconsole start. The patch still contains the extra debugging output showing the time since the last Sky/Tempest message. If the fix works and your consoles have not frozen for a number of days I will release a full new version with this debugging output removed.
I have now patched to 3.1.4. I waited until I had 24 hours on 3.1.2 and my console didn’t get stuck! There was a lost connection at about noon time but the console continued to display the changing data.
Thanks for the updates @padieter and @eric. Glad to hear all is good and the console is not locking up. I’m getting more confident that we have solved the issue, but we’ll need a few more days to be sure. Keep me updated and let me know if it breaks. If not, I can start thinking about rolling the changes we’ve made into a new version.
Thanks for the feedback @eric and @padieter. I think we can be fairly confident that the bug has gone. I don’t think the issue was the secure websocket, so I am going to try one more patch with that re-implemented. I’ll probably have that ready tomorrow/Saturday. If that runs OK as well, then I can get all the changes rolled into a new version. Thanks again for all the help!