commits
Why:
The connection check loop had no delay and would cause a DOS condition for the desktop Lovense Connect app in release builds of the library.
The DOS condition would not happen in debug builds likely due to speed reasons.
Also portig the tests to the new system
This change also corrects the device config JSON schema
I'll note that the fact that the rotational arm on the Ida is effectly
disabled if you're using a pre v3 client, which seems wrong: I get that
oscilator devices are not supported by VibrateCmd's fallback, but single
direction roation seems like it should work (as any such actuators would
have been mapped as vibes in prior versions)
Also adds tests for other WeVibe protocols and a fix for multi-motor devices.
We're going to be on more latent networks with the new mobile
capabilities so up this quite a bit. If people want a lower ping for
timeout, that's what the buttplug protocol ping is for.
If we have no way to continue receiving commands, just leave the event
loop.
Fixes #501
Otherwise things could get VERY weird when we share the object as
as Arc<>. This does possibly confuse lifetimes
Those messages are not in spec v3, should use Sensor now
Otherwise it spams even debug messages
Means we can track library compatibility now.
Only run connection checks once every 5 seconds or so.
Affects #479
Why:
The connection check loop had no delay and would cause a DOS condition for the desktop Lovense Connect app in release builds of the library.
The DOS condition would not happen in debug builds likely due to speed reasons.
This change also corrects the device config JSON schema
I'll note that the fact that the rotational arm on the Ida is effectly
disabled if you're using a pre v3 client, which seems wrong: I get that
oscilator devices are not supported by VibrateCmd's fallback, but single
direction roation seems like it should work (as any such actuators would
have been mapped as vibes in prior versions)