This commit contains the infrastructure needed for the
NotificationHandler to send music state information to the device. That
is, it introduces a call onSetMusicState(MusicStateSpec stateSpec), that
in turn sets up an intent to the service, which will then call the
encodeSetMusicState() function of the device. encodeSetMusicState is
available for pebble only. There are empty stubs for other devices.
- dynamically toggle hr sleep support when preference changes
- check hr support dynaically after device info is available to avoid false error message
Also supports double firmware upgrade for Mi1S.
- so far, only hr firmware upgrade is tested for 1S
- adds junit testcases for firmware recognition and handling
This avoids a lot of problems because java
- does not know unsigned values
- jvm and dalvic do not internally support byte and short
- sqlite does not know them either
- listen to notifications early -- the band then actually tells us that
authentication is required
- check for this after sending user info
- add authentication states to GBDevice
- workaround for event problems in pairing activity (delivered although
already unregistered)
- BtLEQueue now deals with gatt events coming *before* connectGatt()
actually returned (namely the connection event)
Added logic to deal with pebble health sleep data.
Added database helper to change the type of a range of samples (needed for sleep data).
Fixes to the Pebble Health sample provider.
Created a new device-independent class ActivityUser to hold the data
Moved the constants from the miband constant class to the ActivityUser class
Removed the miband-specific in favor of common-prefixed preferences (with upgrade support for legacy values)
Changed the way the gender is stored to an integer value
Removed the hardcoded default values for user data in favor of static fields of the ActivityUser class
(i.e. bluetooth-level pairing). Previously the pairing appeared to
happen automagically, but this doesn't work (anymore). So now we first
pair on the bluetooth-level, then application level.
- created and provided by DeviceHelper
- passed from UI to service
- without UI, service uses DeviceHelper directly
=> Cleaner and less duplicated code
I hope I didn't break firmware upgrades on some Mi 1 models
other than mine (my hardware revision is 2).
Upgrades for Mi 1S are currently disabled, we need some brave
souls who can help us test this.
Closes#173
Also see: #169
This still does not fix pbw isntallation problems with FW 3.x since the content does not get copied to the pbw cache yet when content providers are involved
The primary external dir is often a user-partition on the internal
storage medium. This one is safe in multi-user environments.
The "removable storage" (sdcard) on the other hand can be read by
everyone. If the former is not available or not writable, use the
latter.
closes#153
It should fix installation of all .pbl files (tested with Spanish, German, French and the inofficial Japanese language pack)
This also marks the return of the infamous STM32CRC, this time with hopefully fixed tail code.
NOTES:
- YOU SHOULD NOT TRY THIS YET ;)
- This was only tested with the unoffical japansese language pack
- Problably needs proper crc calculation (I just hardcoded the one for the japanese language pack)
The notfification APIs now use NotificationSpec as their only parameter, which
contains all information (required and optional ones).
We no longer have separate methods and actions for SMS/EMAIL/GENERIC anymore.
The type of notification is important now, not how we received them technically.
Fixes#102
We could also use a whitelist approach instead, but we can't be sure that the user installs a file with .fw extension, hence i preferred to only avoid loading files which are known to be of the wrong type.
Previously, the DeviceCommunicationService was invoked directly,
via
Intent intent = new Intent(foo, bar);
intent.setExtra(EXTRA_BAZ, baz);
startService(...);
and this was scattered throughout GadgetBridge.
Now there is a "frontend" available, so that you can call
the service more easily, like
GBApplication.deviceService().connect();
For a start, this client interface (DeviceService) actually
implements the same interface (EventHandler) as the receiving side
(DeviceSupport). This may change in the future.
This will also make testing much easier, because we can use
this client interface to invoke the test service as well.
- make FwAppInstallerActivity wait for a completely initialized device
- check basalt/aplite compatibility with pbw to be installed and report intead of crashing
- fix crash when trying to install pbw with all app slots full
- version information is now provided implicitly by device initialization
- ACTION_REQUEST_VERSIONINFO is now ACTION_REQUEST_DEVICEINFO and it will
return the current device state of the service without asking any DeviceSupport
instance.
- ACTION_CONNECT now implicitly answers with a device update intent if it
IS already connected.
- the BASE_UUID is shared between all BTLE devices, not miband specific. So are the UUID DESCRIPTORS. Hence these have been moved to AbstractBTLEDeviceSupport
- the gatt-characteristic-property-constants used previously in MiBandNotifyAction are also general, and exposed by the gatt subsystem, hence the specific action has been dropped moving the logic to the NotifyAction class
- the logic for checking the gatt-characteristic-property-constants has been extended also to the ReadAction and WriteAction class, this way we won't try to read (or write) a characteristic that isn't readable (or writeable)
- model package contains mostly shared interfaces (UI+service), not named GB*
- impl package contains implementations of those interfaces, named GB*
the impl classes should not be used by the service (not completely done)
- the service classes should mostly use classes inside the service and deviceevents
packages (tbd)
Every device now has two packages:
- devices/[device name] for UI related functionality
- service[device name] for lowlevel communication
