I’m currently testing the Speedtest integration and, so far, no issues!
However, I find that the display in the interface might be a bit too detailed for some use cases. It would be interesting to be able to choose between three display levels:
Full mode: all available information (upload and download speeds, ping, server, etc.).
Intermediate mode: three boxes displaying upload and download speeds as well as ping, accompanied by the graph to visualize performance evolution and possibly identify outages, if the data allows.
Compact mode: only upload and download speeds and ping, for a more discreet display on the dashboard.
This would allow everyone to adapt the widget to their needs and the available space on their dashboard, while still providing access to detailed information for those who want it.
In any case, thanks for the work done on this integration!
If I understand correctly, there are two types of people in the game: guys like you, who work on integrations, and guys like me, who show up with all sorts of weird requests!
And on top of that, if I understand correctly, we might even need to slow you down a bit in your development pace, because at this rate, we don’t even have time to finish testing one integration before three new ones arrive!
In any case, hats off for the work and thanks for pushing Gladys forward so quickly!
Speeds, ping and graph: three boxes (download, upload, ping) and the graph;
Speeds and ping: just the three boxes.
The two short versions no longer have the button: a test is always launched from the integration configuration. Update from Gladys and let me know how it goes!
Does the graph allow detecting an Internet outage, or are the data smoothed out, making an interruption not clearly visible?
2°) Using data in scenes
Is it possible to use the test results to trigger actions in Gladys? For example, if the ping exceeds 300 ms, send a notification, then run a new test two minutes later to verify that it’s not an outlier.
If the issue is confirmed, we could then consider taking further action, such as restarting the box.
The idea would be to set up automatic monitoring of the Internet connection, with a verification before triggering an action, to avoid false alerts and unnecessary restarts.
Well spotted @Damien! That was intentional: for the short modes, I chose to remove the button (and the date) to keep a really compact widget. But since everyone has their own use case, 1.6.0 adds a setting per widget, Button « Run a test »: as before, always, or never. You can put it back on your reduced widget
Outages: honestly, the graph isn’t designed for that. Tests run every hour by default (minimum 10 minutes), a failed test sends no value, and the graph connects the dots: an outage between two tests isn’t visible.
Scenes: this is in place with 1.6.0:
the trigger Automatic Speed Test Failed, which runs when an automatic test cannot complete (often an outage);
the action Run a Speed Test, which waits for the result and passes it to the next actions.
For your example: trigger « device value » ping > 300 ms → notification → wait 2 minutes → Run a Speed Test → « Continue only if » the ping is still above 300 ms → the rest.
To restart the box, you’ll need the integration of your box: which one do you have?
Personally, I have an SFR box and the restart button directly in the integration is really handy! No more running to the garage to press the box’s button.
And thanks for the scene diagram, it really helps to understand how to set everything up!
Super And good news: the SFR Box integration also has a scene action Restart the SFR Box. You can therefore place it at the end of your scene, after the “Continue only if”, to complete the loop.
Small tip: the box takes a few minutes to come back after a restart, and during this time the ping remains poor. Keep the wait and the second test before the restart, to avoid restarting it in a loop.
I’ll check it out
