b99b8522e02c8e59af0a2ca9beec55f8e1e0b8f4: Bug 1368418 part 3 - Remove nsCategoryImp. r?emk draft
Xidorn Quan <me@upsuper.org> - Mon, 29 May 2017 16:17:39 +1000 - rev 586194
Push 61330 by xquan@mozilla.com at Mon, 29 May 2017 23:08:38 +0000
Bug 1368418 part 3 - Remove nsCategoryImp. r?emk MozReview-Commit-ID: 5qCoeqfM2s5
dd504d0ed98bce5aaab663f4b5da4d021334632e: Bug 1368418 part 3 - Remove nsCategoryImp. r?emk draft
Xidorn Quan <me@upsuper.org> - Mon, 29 May 2017 16:17:39 +1000 - rev 586193
Push 61329 by xquan@mozilla.com at Mon, 29 May 2017 23:04:07 +0000
Bug 1368418 part 3 - Remove nsCategoryImp. r?emk MozReview-Commit-ID: 5qCoeqfM2s5
07772bd9728f4a704b67bc874e26e15ef785bc11: Bug 1345898: P3. Flush decoder on Android. r?jolin draft
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 30 May 2017 00:31:11 +0200 - rev 586192
Push 61328 by bmo:jyavenard@mozilla.com at Mon, 29 May 2017 22:53:43 +0000
Bug 1345898: P3. Flush decoder on Android. r?jolin In the situation described in P2, on android where the decoder is recycled rather than shut down, we would have potentially returned multiple frames with the same timestamp. It is necessary to flush the decoder. MozReview-Commit-ID: 6ovaD9PEdOG
53b17359820617db47006588ccd71ea2aa1b5937: Bug 1107364 - remove findNodeByDetails and add a hash to find a node by details speedily, r?mak draft
milindl <i.milind.luthra@gmail.com> - Tue, 30 May 2017 01:21:25 +0530 - rev 586191
Push 61327 by bmo:i.milind.luthra@gmail.com at Mon, 29 May 2017 22:49:08 +0000
Bug 1107364 - remove findNodeByDetails and add a hash to find a node by details speedily, r?mak This patch removes the interface/header/implementation if `findNodeByDetails` and replaces it by a map [{uri, itemId, time} -> Node] which can recover a node speedily by details. MozReview-Commit-ID: LaO9DH2J5mK
ae3d5b425ce894bdab1ff7d8e574e9ae3b58596b: Bug 1361664 - Part2. Create MVP Triggers and Deeplinks for Mobile Marketing Automatin. r?maliu draft
Nevin Chen <cnevinchen@gmail.com> - Wed, 17 May 2017 01:25:45 +0800 - rev 586190
Push 61326 by bmo:cnevinchen@gmail.com at Mon, 29 May 2017 22:42:47 +0000
Bug 1361664 - Part2. Create MVP Triggers and Deeplinks for Mobile Marketing Automatin. r?maliu This patch do 6 things. They are related so I put them in the same patch. 1. Extract MmaEvent Name 2. If MMA is diabled, don't send event. 3. Add check before sending 'Set Default Borwser' deep link 4. Add user attribute for delay messaging focus install status. 5. If the user pref off at runtime, we ask Leanplum to stop and prevent our app sending event to Leanplum. 6. Fix proguard. Only keep necceary files. MozReview-Commit-ID: APEmr1JXBLH
248484ac438ec47fc7b91048c0ffaa8d6a559ad3: Bug 1361664 - Part1. Udate MMA API, make init() accept activity. r?maliu draft
Nevin Chen <cnevinchen@gmail.com> - Wed, 24 May 2017 11:23:10 +0800 - rev 586189
Push 61326 by bmo:cnevinchen@gmail.com at Mon, 29 May 2017 22:42:47 +0000
Bug 1361664 - Part1. Udate MMA API, make init() accept activity. r?maliu MozReview-Commit-ID: AcfoeCPV8SV
d92cbc57d361db19f7a45fd197b924302738bb4a: Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 18:24:29 -0400 - rev 586188
Push 61325 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:33:46 +0000
Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm MozReview-Commit-ID: EROfGuYQ6a0
22da4f664e2ce17ea956da66087e4e78c90b5592: Bug 1367591 - Process gDelayedAnnotations on Windows and Mac as well, not just on Linux. r?jdm draft
Markus Stange <mstange@themasta.com> - Sun, 28 May 2017 22:30:22 -0400 - rev 586187
Push 61325 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:33:46 +0000
Bug 1367591 - Process gDelayedAnnotations on Windows and Mac as well, not just on Linux. r?jdm MozReview-Commit-ID: LQrU8NTdj7V
e16509ff11a44db136e82fc6068a28f4a55e4a6e: Bug 1345898: P3. Flush decoder on Android. r?jolin draft
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 30 May 2017 00:31:11 +0200 - rev 586186
Push 61324 by bmo:jyavenard@mozilla.com at Mon, 29 May 2017 22:31:56 +0000
Bug 1345898: P3. Flush decoder on Android. r?jolin In the situation described in P2, on android where the decoder is recycled rather than shut down, we would have potentially returned multiple frames with the same timestamp. It is necessary to flush the decoder. MozReview-Commit-ID: 6ovaD9PEdOG
1a06cd051fc6544868d9d5980c13b327756e9afb: Bug 1345898: P2. Flush rather than drain decoder. r?gerald draft
Jean-Yves Avenard <jyavenard@mozilla.com> - Tue, 30 May 2017 00:30:54 +0200 - rev 586185
Push 61324 by bmo:jyavenard@mozilla.com at Mon, 29 May 2017 22:31:56 +0000
Bug 1345898: P2. Flush rather than drain decoder. r?gerald When encountering a change of stream ID while performing an internal seek to resume from the last playback position (following WAITING_FOR_DATA), we may as well perform a flush rather than a drain as the decoded data won't be used anyway. MozReview-Commit-ID: E7wgTndTThs
cfa688b52c7af1c6cafc8c9636d62067f7b98b88: Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 18:24:29 -0400 - rev 586184
Push 61323 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:24:53 +0000
Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm MozReview-Commit-ID: EROfGuYQ6a0
4752ac1287f384fcae6f1bffd7bc174c4769705b: Bug 1365637 place WE sidebars into the photon sidebar dropdown, r?mattw,Gijs draft
Shane Caraveo <scaraveo@mozilla.com> - Mon, 29 May 2017 15:20:48 -0700 - rev 586183
Push 61322 by mixedpuppy@gmail.com at Mon, 29 May 2017 22:21:08 +0000
Bug 1365637 place WE sidebars into the photon sidebar dropdown, r?mattw,Gijs MozReview-Commit-ID: 3ZM9mXjEbWh
74e4662e2d9747eac7965db7b3ef376fe682401b: Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 18:11:01 -0400 - rev 586182
Push 61321 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:17:06 +0000
Bug 1330185 - Use a top-level PProfiler protocol to control the profiler in other processes. r=njn, r=billm MozReview-Commit-ID: EROfGuYQ6a0
b4ff8e1f6913fc9fdf32a74c8474a84fc1d76c01: Bug 1330184 - Remove main-thread-only assertions. r=njn draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 15:16:34 -0400 - rev 586181
Push 61320 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:16:19 +0000
Bug 1330184 - Remove main-thread-only assertions. r=njn I think all of these assertions are now unnecessary. MozReview-Commit-ID: 9EI195QsizN
69eebd4852fd8cba3d24e4ad5e5bd274d1f02917: Bug 1330184 - Register/unregister the IOInterposeObserver on the main thread, regardless of what thread the profiler is started / stopped on. r=njn draft
Markus Stange <mstange@themasta.com> - Tue, 16 May 2017 17:35:05 -0400 - rev 586180
Push 61320 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:16:19 +0000
Bug 1330184 - Register/unregister the IOInterposeObserver on the main thread, regardless of what thread the profiler is started / stopped on. r=njn MozReview-Commit-ID: 8Y0rspxBJw3
7d1e50d812f7718063eacd4514cd88de3a1d93d2: Bug 1330184 - Allow StreamMetaObject to be called on a background thread, but only include startTime and version for those calls. r=njn draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 13:44:28 -0400 - rev 586179
Push 61320 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:16:19 +0000
Bug 1330184 - Allow StreamMetaObject to be called on a background thread, but only include startTime and version for those calls. r=njn MozReview-Commit-ID: 3AuX0a2Brrd
e4bbc5990383b5fbd699a29c9886a40f5e6f6341: Bug 1330184 - Allow notifying observers for profiler state changes on background threads. r=njn draft
Markus Stange <mstange@themasta.com> - Mon, 29 May 2017 17:12:50 -0400 - rev 586178
Push 61320 by bmo:mstange@themasta.com at Mon, 29 May 2017 22:16:19 +0000
Bug 1330184 - Allow notifying observers for profiler state changes on background threads. r=njn MozReview-Commit-ID: GlkVwGTa2b4
3c507afb6d17b7f40620ad205cb945db1bab6185: Bug 1367488 - Pre-render offscreen portions of scrollbar thumbs inside an iframe. r=mstange draft
Botond Ballo <botond@mozilla.com> - Mon, 29 May 2017 18:05:05 -0400 - rev 586177
Push 61319 by bballo@mozilla.com at Mon, 29 May 2017 22:15:27 +0000
Bug 1367488 - Pre-render offscreen portions of scrollbar thumbs inside an iframe. r=mstange MozReview-Commit-ID: LCBHnFJdGtp
8fae9b082c76db220a0a5a61d9ea76a79063c847: Bug 1345898: P3. Flush decoder on Android. r?jolin draft
Jean-Yves Avenard <jyavenard@mozilla.com> - Mon, 29 May 2017 23:48:44 +0200 - rev 586176
Push 61318 by bmo:jyavenard@mozilla.com at Mon, 29 May 2017 21:51:32 +0000
Bug 1345898: P3. Flush decoder on Android. r?jolin In the situation described in P2, on android where the decoder is recycled rather than shut down, we would have potentially returned multiple frames with the same timestamp. It is necessary to flush the decoder. MozReview-Commit-ID: 6ovaD9PEdOG
95c777f179520ae86430796446572ca08698a0db: Bug 1345898: P2. Flush rather than drain decoder. r?gerald draft
Jean-Yves Avenard <jyavenard@mozilla.com> - Mon, 29 May 2017 23:44:28 +0200 - rev 586175
Push 61318 by bmo:jyavenard@mozilla.com at Mon, 29 May 2017 21:51:32 +0000
Bug 1345898: P2. Flush rather than drain decoder. r?gerald When encountering a change of stream ID while performing an internal seek to resume from the last playback position (following WAITING_FOR_DATA), we may as well perform a flush rather than a drain as the decoded data won't be used anyway. MozReview-Commit-ID: E7wgTndTThs
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 tip