Merge lp:~smspillaz/unity/unity.less-paint-insanity into lp:unity
Status: | Superseded |
---|---|
Proposed branch: | lp:~smspillaz/unity/unity.less-paint-insanity |
Merge into: | lp:unity |
Diff against target: |
158 lines (+31/-72) 2 files modified
plugins/unityshell/src/unityshell.cpp (+31/-68) plugins/unityshell/src/unityshell.h (+0/-4) |
To merge this branch: | bzr merge lp:~smspillaz/unity/unity.less-paint-insanity |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Daniel van Vugt | Needs Fixing | ||
Jason Smith | Pending | ||
Marco Trevisan (Treviño) | Pending | ||
Review via email: mp+109079@code.launchpad.net |
This proposal supersedes a proposal from 2012-06-07.
This proposal has been superseded by a proposal from 2012-06-12.
Description of the change
Removes the whole scanning of the list thing in getWindowPaintList and the whole "trying to figure out where the window was because we don't actually know" in glDraw by scanning the paint list again and again.
There's a much better way of doing this - just hook glPaint, skip the wrap chain to core (eg every single plugin skipped) and tell core not to do anything with the window.
I have no idea why there is a bunch of logic about whether or not to paint the panel in glDraw (that was moved to glPaint). Someone enlighten me?
Unmerged revisions
- 2405. By Sam Spilsbury
-
Remove unused method and rename methods to make more sense
- 2404. By Sam Spilsbury
-
Merge lp:unity
- 2403. By Sam Spilsbury
-
Basic documentation, not really a substitute for better code ...
- 2402. By Sam Spilsbury
-
Nuke FBOBindingQuery
Interface - 2401. By Sam Spilsbury
-
unsigned int -> enum && move namespace compiz {} to namespace unity::compiz
- 2400. By Sam Spilsbury
-
Change namespace to unity::compiz
- 2399. By Sam Spilsbury
-
Merge lp:unity
- 2398. By Sam Spilsbury
-
Merge lp:unity
- 2397. By Sam Spilsbury
-
Test prohibited paint requestor retreival
- 2396. By Sam Spilsbury
-
Tests for RepaintPending value
Bump?
I'd like to see this go in, because it should reduce lots of list copying within unity and also lots of unecessary list traversal, which should represent a nice speed boost. In addition, its far simpler in its implementation and far less bug prone.
I'm sure testing is probably an issue. I'll have another quick look at the code and see if I can think of a way to get it under test.