Merge lp://qastaging/~vanvugt/unity/fix-953089 into lp://qastaging/unity
Proposed by
Daniel van Vugt
Status: | Merged |
---|---|
Approved by: | Marco Trevisan (Treviño) |
Approved revision: | no longer in the source branch. |
Merged at revision: | 2128 |
Proposed branch: | lp://qastaging/~vanvugt/unity/fix-953089 |
Merge into: | lp://qastaging/unity |
Diff against target: |
206 lines (+42/-34) 5 files modified
manual-tests/Tapping.txt (+20/-0) plugins/unityshell/src/LauncherController.cpp (+9/-25) plugins/unityshell/src/LauncherController.h (+2/-2) plugins/unityshell/src/unityshell.cpp (+10/-6) plugins/unityshell/src/unityshell.h (+1/-1) |
To merge this branch: | bzr merge lp://qastaging/~vanvugt/unity/fix-953089 |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Marco Trevisan (Treviño) | Approve | ||
Mirco Müller (community) | Approve | ||
Review via email:
|
This proposal supersedes a proposal from 2012-03-16.
Commit message
Use more accurate timing to detect key taps (for Super or Alt).
Description of the change
Use more accurate timing to detect key taps (for Super or Alt). The previous
timing methods were failing on slow or busy systems, causing taps to be
misdetected and ignored. Instead use the timestamps that come with each
keypress/release event. (LP: #953089)
To post a comment you must log in.
Code looks good, tested it myself by loading up my CPU with compiling-tasks in the bg... the response from <Super> and <Alt> feel solid, but a manual-test would be needed to help guide the QA-people be able to verify this fix. I'd say one can "simulate" a busy system with doing a CPU-hogging "make -j4" of e.g. nux or unity in the background, while running unity itself and testing the response-behaviour for <Super> and <Alt>.
With a manual-test in place, I'm gladly marking this as approved.
See the text-files under unity/manual-test to get an idea, what a manual-test is made up of.