Currently the tests do little more than fire up the simulator and
switch into (and out of) the built in applications. However this is
useful and allows us to fully integrate as a CI job.
Unfortunately the numpy warning from pysdl2 mean we have been forced
to disable all warnings to prevent pytest from collecting and reporting
them.
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
Currently boardless targets including softdevice, sudmodules and sim
cannot be run unless a dummy value of BOARD is supplied. Fix this by
distinguishing between conditional and unconditional expansions of the
BOARD variable.
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
wasp-os is starting to developer proper support for multiple boards
(currently other boards such as nitrogen and ds-d6 test builds and
proof-of-concept stuff) then we need to start placing binaries for each
device into seperate directories (other wise creating a binary distro
won't work).
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
If the preprocessor fails for any reason then an empty or partially
processed watch.py file is created which will satisfy the dependency
checks in any later invocation of make which can silently break the
build. Cleaning up on failure prevents this from happening.
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
Generate a .zip file that works with SoftDevice 2.0.1 (still looking
for the recipe that works with 5.0), fix up the bootloader to work with
DaFlasher and add an extra feature in micropython to cope better.
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
The paths passed to the preprocessor are currently wrong. Fix them!
Fixes: 64afca1 ("Makefile: fix make sim (allow BOARD to be unset)")
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
The preprocess to create watch.py was hard coded for PineTime so changed to $(BOARD)
This will be required for other watches like the P8
Signed-off-by: Steve Amor <steveamor@users.noreply.github.com>
Currently when the the bootloader is modified if fails to build because
the wildcard in the hex merge picks up both the old and new binaries.
Fix this by explicitly deleting the old one.
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
Currently there's no fancy algorithms to estimate stride length. Just
pure simple step counting directly from the hardware's "intelligence
engine".
Signed-off-by: Daniel Thompson <daniel@redfelineninja.org.uk>
The micropython build system has incomplete dependancy tracking, especially
with regard to frozen manifests. Work round this by nuking the frozen
content before attempting a build. Sure this is a hack... but much
better than developing a "make clean; make all" habit.
This not only avoids the use of a deprecated feature but it also
makes it easier to decouple the wasp/ directory structure from
what actually gets copied to the target.
ml server