f5a569610f
Wrap DeviceSupport instances.
...
The wrapper provides support for busy-checking and throttling
(sometimes I get multiple events of the same kind within milli seconds
and the Mi Band vibrates 20 times)
2015-06-06 23:54:34 +02:00
f6d5767276
Remove reference to PebbleIOThread from BluetoothCommunicationService
...
Let PebbleSupport do this instead.
2015-06-06 23:24:00 +02:00
3fe9195d0d
Avoid warning in unsetDynamicState()
2015-06-06 23:13:26 +02:00
020d8d74d6
Added "Fetch Activity Data" to the context menu.
...
With some kind of progress reporting during the fetching #45
2015-06-06 19:39:04 +02:00
9e4e50be47
Initial work on synchronizing activity data with feedback.
...
A device now has a busy flag (set during synchronization). While busy,
no other communication with the device shall occur (TODO)
Refactors the non-bluetooth actions a bit #45
Next step: make use of the busy state in ControlCenter (show
a busy cursor) and in BluetoothCommunicationService (to not call other
operations while busy)
2015-06-06 00:40:16 +02:00
2f0d00d645
avoid WITHOUT ROWID when not running at least lollipop #62
2015-06-06 00:10:38 +02:00
1e89b12b15
Avoid creating lots of SimpleDateFormat instances
2015-06-05 23:09:28 +02:00
dea4ee82a1
Merge branch 'db_refactoring_proposal'
2015-06-05 21:57:55 +02:00
c4096e6d3c
Add daniele's db optimization the WITHOUT ROWID
...
closes #62
2015-06-05 21:46:56 +02:00
b12a3e74cd
MorpheuzSupport: remove limit of 54 samples. Morpheuz 3.1 supports 60
2015-06-05 20:26:11 +02:00
baecc20742
SleepMonitorActivity: Rotate annotation text for better readability
2015-06-05 19:59:58 +02:00
813a02d5c7
Unify SleepMonitor with code from PR #59 . Thanks Daniele!
2015-06-05 19:32:09 +02:00
2f1908e480
Make sure to always close the database
2015-06-04 23:45:46 +02:00
fc374881c5
Get rid of the unused ID in the DB, this could help in:
...
* reducing the used space
* reducing the time needed for inserts
2015-06-04 21:37:48 +02:00
7ad38c5e1a
Merge pull request #61 from danielegobbetti/fix_data_in_the_future
...
Do not iterate over the whole buffer, but only on the effectively use…
2015-06-04 21:15:52 +02:00
a1ff9aab21
Do not iterate over the whole buffer, but only on the effectively used part.
...
This should fix the "data in the future" issue
2015-06-04 18:56:35 +02:00
7d86396e30
Support CM 12.1 Music App
2015-06-02 21:48:21 +02:00
32b2500d6b
Merge pull request #60 from danielegobbetti/fix_data_in_the_future
...
Do not allocate the buffer if there's no data available.
2015-06-02 21:43:32 +02:00
406f9ab90d
Do not allocate the buffer if there's no data available.
...
Perhaps could fix the "data in the future" issue
2015-06-02 18:33:12 +02:00
9e2d32c33f
Fixed timestamp conversion for DB, added local variables to make the code clearer
2015-06-01 22:20:28 +02:00
ea97a902d1
*unchecked* Storing activity values in the DB.
2015-06-01 22:20:28 +02:00
8b268a676c
Refactoring of the data ingestion method.
...
* no more ByteBuffer, but a fixed size byte array that gets flushed everytime it's needed
* log of activity data to a separate file (no DB integration yet)
* the size of the buffer must be a multiple of 3 (1 minute of data = 3 bytes) and 20 (the normal size of the chunks we get from the device)
* better logging and more comments in code
2015-06-01 22:20:22 +02:00
5d950dc407
Fix reconnection after connection loss
...
Well, obviously we must not ignore connection state changes even if they
come with an error code.
Unfortunately the API docs are a bit terse in that respect.
2015-06-01 21:47:02 +02:00
7f89f4bb57
Some fixes regarding device initialization, should avoid disconnects
...
Avoid repeated initializations and device info requests. Fix unsetting
of dynamic state (e.g. battery info) when initialized.
2015-06-01 21:47:02 +02:00
f60903699e
Delete MiBandSupport.java.orig
...
I believe it was committed by mistake (perhaps a merge conflict?) It's out of date in any case.
2015-06-01 21:47:02 +02:00
11884d8073
Improve logging
2015-06-01 21:47:02 +02:00
75b9fe4c4d
ignore connection state changes with status != GATT_SUCCESS
...
And add logging. Maybe this helps with #57
2015-06-01 21:47:02 +02:00
228e922ce7
forgotten refactor of a variable
2015-06-01 21:47:01 +02:00
e4076dc725
Ingesting the activity data the right way, but they are still being thrown away afterwards.
2015-06-01 21:47:01 +02:00
92caed5af4
Some cleanup
2015-06-01 21:47:01 +02:00
a7792f6b72
improve some log statements
2015-06-01 21:47:01 +02:00
637b43e892
Also wait for descriptor reads and writes
...
(not just characteristic reads/writes)
This fixes initialization of notification characteristics
(activity, sensor data, battery, ...)
2015-06-01 21:47:01 +02:00
f004b7b11c
WIP for fetching activity data
...
This probably affects #44 and #45
Thanks go to Daniele Gobbetti <daniele@gobbetti.name>!
2015-06-01 21:47:01 +02:00
6ea9537d38
Start sleep activity with context menu (long press on a device)
...
This allows to see sleep data even if devices are not connected.
2015-06-01 16:57:45 +02:00
81b1d1d28d
Pebble: Put WeatherNeat test in separate file
2015-05-30 22:24:53 +02:00
603d31a59e
add some constants to GBActivitySample
2015-05-30 21:52:29 +02:00
41207516b1
Move BluetoothChangeReceiver to externalevents
2015-05-30 21:40:42 +02:00
900b3f3833
use ORDER BY timestamp in SQL statement
...
Also optimize imports...
2015-05-30 17:40:23 +02:00
e79f4523c3
Add SQLite database support for activity data
...
This is now used for morpheuz data and the SleepMonitorActivity
2015-05-30 17:28:03 +02:00
08fbbb9152
SleepMonitor: draw colored bars instead of a graph
...
red = awake
yellow = light sleep
greep = deep sleep
2015-05-27 00:59:05 +02:00
06ee7efe79
update CHANGELOG.md
2015-05-25 01:32:16 +02:00
8366af736c
Pebble: Fix firmware installation when all 8 app slots are in use
...
We were trying to get a free slot when installing a firmware which is not neccessary.
This should also fix firmware installation in recovery (#54 )
2015-05-25 01:26:27 +02:00
6fb6b5c164
Merge pull request #55 from danielegobbetti/master
...
Use constants instead of hardcoded values
2015-05-24 23:16:52 +02:00
8a26ce9d67
Pull evaluateGBDeviceCommand() and handle* methods up
...
Otherwise, MiSupport can't use 'em
2015-05-23 00:54:14 +02:00
4518e8819d
Make file-logging configurable
2015-05-23 00:45:12 +02:00
c469248de1
Updated from transifex (thanks!)
2015-05-22 23:21:37 +02:00
880dc7b3a4
Improve BLE gatt handling
...
- close() the gatt instance when explicitly disconnecting
- create a new gatt instance when explicitly connecting
Unfortunately I still appear to get spontaneous disconnects after some
notifications.
2015-05-22 23:18:29 +02:00
f54927624b
Move non Pebble specific code from PebbleIoThread to AbstractBTDeviceSupport
2015-05-22 20:20:33 +02:00
8309234784
Pebble: fix wrong calculation of timestamp.
2015-05-21 18:57:34 +02:00
7e2545f9b4
Use constants instead of hardcoded values
2015-05-21 18:19:07 +02:00
cpfeiffer