diff options
| author | Charles Munger <clm@google.com> | 2019-11-12 09:00:48 -0800 |
|---|---|---|
| committer | Charles Munger <clm@google.com> | 2019-11-15 10:17:54 -0800 |
| commit | 80b1c4dcc8c5c9494c98693100a782644a16042d (patch) | |
| tree | fc37e55efa895f91ca06927a8b479cbca5014272 /tools/aapt2/java/JavaClassGenerator_test.cpp | |
| parent | 1695301cc944259d6df9589b6b51189d145dd62d (diff) | |
Deprecate implicit Handler constructors
These have been the cause of bugs in the past, such as b/10150258,
because it's not always obvious or guaranteed that a calling thread will
always have a looper, or that it's actually the looper you want.
Specifying the looper explicitly, even if it's just Looper.myLooper(),
makes the code clearer. Since callback APIs generally should have
Executor variants added, it might be nice to also deprecate the
constructors or options that rely on the calling thread's looper, but
since I don't have an inventory of all those I'm starting with this.
Bug: 144042891
Test: No behavior changes
Change-Id: I0e7ba80d6b89a866ea715c790dd602ef728bf5bd
Diffstat (limited to 'tools/aapt2/java/JavaClassGenerator_test.cpp')
0 files changed, 0 insertions, 0 deletions