Merge lp:~alan-griffiths/miral/hide-streams-from-Window-Management-API into lp:miral
Proposed by
Alan Griffiths
on 2016-09-01
| Status: | Merged |
|---|---|
| Approved by: | Gerry Boland on 2016-09-05 |
| Approved revision: | 310 |
| Merged at revision: | 313 |
| Proposed branch: | lp:~alan-griffiths/miral/hide-streams-from-Window-Management-API |
| Merge into: | lp:miral |
| Diff against target: |
324 lines (+16/-160) 7 files modified
debian/libmiral1.symbols (+0/-2) include/miral/stream_specification.h (+0/-42) include/miral/window_specification.h (+7/-5) miral/CMakeLists.txt (+0/-1) miral/basic_window_manager.cpp (+0/-25) miral/symbols.map (+0/-1) miral/window_specification.cpp (+9/-84) |
| To merge this branch: | bzr merge lp:~alan-griffiths/miral/hide-streams-from-Window-Management-API |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Gerry Boland | 2016-09-01 | Approve on 2016-09-05 | |
|
Review via email:
|
|||
Commit Message
Don't expose BufferStreams to window management when creating or modifying windows.
To post a comment you must log in.
| Chris Halse Rogers (raof) wrote : | # |
| Alan Griffiths (alan-griffiths) wrote : | # |
> It's not clear why you're removing pixel_format(), buffer_usage(), and
> content_id() at the same time. Is that deliberate?
Well, it could have been a separate MP, but they are also properties that WM shouldn't be (and AFAICS isn't) messing with.
| Chris Halse Rogers (raof) wrote : | # |
Hm. My understanding is that those *are* things the shell might want to mess with or, rather, to deny creating a window with certain properties.
So the WM doesn't need to deal with them, no, but something in MirAL might?
lp:~alan-griffiths/miral/hide-streams-from-Window-Management-API
updated
on 2016-09-05
- 310. By Alan Griffiths on 2016-09-05
-
Remove dead file from CMakeLists.txt
| Gerry Boland (gerboland) wrote : | # |
I don't need this yet - perhaps I never will. Ok to remove
review:
Approve

It's not clear why you're removing pixel_format(), buffer_usage(), and content_id() at the same time. Is that deliberate?