Fixed
Status Update
No update yet.
Comments
yb...@google.com
Which device is issue being reported on? Can you get any logs from the device for us after you turn on verbose logging ?
[Deleted User] <[Deleted User]>
A few things in addition to the questions in #2:
1. Can you verify that this problem still occurs in beta03? This is available now.
2. Can you check your merged manifest has a network permission (ACCESS_NETWORK_STATE)? It should get merged in automatically.
1. Can you verify that this problem still occurs in beta03? This is available now.
2. Can you check your merged manifest has a network permission (ACCESS_NETWORK_STATE)? It should get merged in automatically.
yb...@google.com
#3: have now tried beta03 and it's the same behaviour... periodic work schedule breaks down when network constraint is applied (after a period of time with no network). And yes, ACCESS_NETWORK_STATE is in the merged manifest... it's actually in the main manifest anyway.
#2: I'll try...
#2: I'll try...
[Deleted User] <[Deleted User]>
Re #2 getting logs... I have tried previously to get logs from the user, and on each occasion the logs started *after* the point of interest, as if the logs had been cleared just before (e.g. just after a problem), so they've not been of any use.
The way I'm getting logs is via a button in the widget configuration activity which runs the following:
Runtime.getRuntime().exec("logcat -f " + logFile + " *:V");
Is there a better way of getting logs from a user (where the user is not an advanced user and can only just about handle pressing a button and sending a text file)?
The way I'm getting logs is via a button in the widget configuration activity which runs the following:
Runtime.getRuntime().exec("logcat -f " + logFile + " *:V");
Is there a better way of getting logs from a user (where the user is not an advanced user and can only just about handle pressing a button and sending a text file)?
yb...@google.com
Looks like this issue isn't specific to Unique**Periodic**Work ... with a network constraint applied to UniqueWork, if it is scheduled for a time when there is no network, it doesn't go off, even when network is restored. Again, on this user's Android 4.3 device... on the devices I've tested, it works fine.
Do I need to do anything special to make this work on Android 4.3?
Do I need to do anything special to make this work on Android 4.3?
ap...@google.com
And periodic work is running just fine on the problem device without the network constraint applied... runs like clockwork.
Description
Version used: 2.0.0
Devices/Android versions reproduced on: N/A
In current version of room if I mark method in DAO with @Transaction, and if it uses any disposable e.g. Single, the generated code will throw main thread access exception at runtime.
I would expect one of two behaviours instead:
1. It would be ideal if room could somehow defer start/end transaction to match thread that is used on Disposable, but I can imagine that this can be not possible.
2. To get compile time warning or error that this may be not the correct way of using @Transaction with mix of asynchronous code.
Example of code:
Method in abstract DAO:
@Transaction
open fun upsert(obj: T): Single<Boolean> {
return insert(obj)
.map { true }
.onErrorResumeNext {
if (it is SQLiteConstraintException)
{ update(obj)
.map { true }
} else {
Single.error(it) }
}
.subscribeOn(Schedulers.io()) }
Generated method:
@Override
public Single<Boolean> upsert(Supplier obj) {
__db.beginTransaction();
try {
Single<Boolean> _result = super.upsert(obj);
__db.setTransactionSuccessful();
return _result;
} finally {
__db.endTransaction();
}
}