When I added more parsing of the dive data of the Cosmiq+ in commit
4dff291a1a53 ("Deepblu Cosmiq+: fill in some parsing details") I got the
gasmix units completely wrong and clearly never tested it.
The DC_FIELD_GASMIX reporting uses floating point percentages, not
integer percentages, and instead of reporting 21% as 0.21, we used to
report it as 21.0. It all looked fine in my profiles, because I'd only
tested simulated air dives, and subsurface defaults to air even if
somebody reports crazy impossible gases.
Easy enough to fix, and now actually tested by doing a simulated nitrox
dive.
Reported-by: Michael Werle <mwerle@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This adds support for the different modes (scuba/gauge/freedive) and
teaches the parser to set the right sample interval and get the dive
time right.
In freedive mode, the sample interval is 1s, and the dive time is in
seconds too. In the other modes, the sample interval is 20s, and the
divetime is in minutes.
Also set the right gas for scuba.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This at least approximates downloading a dive from the Deepblu Cosmiq+,
and gives reasonable profiles for my test-dives (in a dive computer
chamber, not real dives).
I'm sure there's a lot to be improved here, and this literally only gets
depth and water temperature, but that seems to be what the Cosmiq+ gives
us.
Lots of credit to Michael Werle for bluetooth packet captures, and some basic analysis of the protocol.
Packet-logging-by: Michael Werle <mwerle@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This does nothing at all, but it adds all the core skeleton
infrastructure for the Deepblu Cosmiq+ dive computer.
Let's see if I can make sense of things and make it download anything.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
