Description:
The "remove" event:
- stops gpsd from polling the device
- does not un-stash the device from the gpsd internal device slots, nor does it deactive the device

Is this by design?


Test #1
do not disconnect gps device
manually call "add"
manually call "remove"
run gps client
gpsd queries a supposedly "removed" device

Test #2
do not disconnect gps device
manually call "remove"
- gpsd: control socket connect on fd 6
- gpsd: <= control(6): removing /dev/ttyUSB0
- gpsd: closing GPS=/dev/ttyUSB0 (7)
manually call "add"
- gpsd: control socket connect on fd 6
- gpsd: <= control(6): /dev/ttyUSB0 already active

Test #3
disconnect gps device
see lots of failed reads & high cpu
manually call remove
with deivce still disconnected, run a gps client
- gpsd: opening GPS data source at '/dev/ttyUSB0'
- gpsd: device open failed: No such file or directory - retrying read-only
- gpsd: read-only device open failed: No such file or directory


Additional info:
* package version(s)
community/gpsd 2.92-1