Attendees (unique): 10
Tridge doesn’t have enough bandwidth for a screencast.
UTC0705
Merged!
UTC0706
Randy: Was this tested with the change?
Peter: No
Merged!
UTC0709
Andy: Allows for picking the internal compass first, in it exists. This allows for setting the correct orientation.
Andrew: I think this will affect device detection in other users, right?
Andy: It will affect people who have this problematic configuration.
Sid: It will ask them to recalibrate first, because the compass has changed.
A: But people don’t recalibrate.
S: It will block their takeoff with a prearm failure.
A: In the past we added a define in hwdef to stop probing a specific external sensor.
- I’m not comfortable with changing it in a way that may affect other users. Better go with the special define.
Andy: I don’t like this workaround, because it prevents you from realizing that something has been misconfigured.
A: Why is it necessary to redefine GET_I2C_DEVICE, isn’t probing the internal bus first sufficient? - Have a board specific define that probes the internal devices first.
- And users will have to reorder the compasses with the MP tool.
UTC0728
Sid: Addresses the situation where DroneCAN compasses will get confused with each other, if you plug-unlplug too many of them. Replacing compasses in DroneCAN was annoying so it was disabled. But we also need to increase the memory allocation for the amount of compasses tracked, from 3 to 6 now.
Andy: Approved.
R: Would like to see more detailed explanation and more testing.
UTC0738
S: Adds support for dedicated parameters to configure UBlox modules.
- We can now have control over all of the features, like we did before M9s.
- It allows e.g. setting the dynamics profile of the module, granular control of constellations and bands.
- It can be disabled by a define at build time.
Currently it adds 8kB to the build.
R: Would be nice to see a write-up on what testing has been done.
Andy: The legacy driver was quite crusty, but the increased size will be a concern. Also testing will need to be thorough.
S: By default the legacy driver behaviour will be taking place, unless you choose to use the new one.
R: If you put this early in the 4.8, it may be able to be introduced gracefully.
S: Actually the new X20 module will need this new driver.
- The new bands will allow sub-10cm accuracy with sattelite-based corrections.
- Testing is very welcome!
UTC0758
Randy: You can take life in your hands and merge it yourself, Andy.
UTC0801
Andy: The manufacturer has switched to another baro and the original config doesn’t work.
R: Will this break old boards?
Andy: I haven’t found any boards with the original spec.
R: I’ve approved it, but let’s try to confirm that the older boards work.