Merge lp:~alan-griffiths/miral/confine_pointer into lp:miral
| Status: | Merged |
|---|---|
| Approved by: | Daniel d'Andrada on 2016-09-23 |
| Approved revision: | 363 |
| Merged at revision: | 358 |
| Proposed branch: | lp:~alan-griffiths/miral/confine_pointer |
| Merge into: | lp:miral |
| Diff against target: |
537 lines (+230/-101) 11 files modified
debian/libmiral1.symbols (+6/-1) include/miral/detail/mir_forward_compatibility.h (+130/-0) include/miral/window_info.h (+3/-0) include/miral/window_specification.h (+4/-98) miral/CMakeLists.txt (+12/-1) miral/basic_window_manager.cpp (+6/-0) miral/mir_features.h.in (+28/-0) miral/symbols.map (+8/-0) miral/window_info.cpp (+11/-0) miral/window_management_trace.cpp (+2/-1) miral/window_specification.cpp (+20/-0) |
| To merge this branch: | bzr merge lp:~alan-griffiths/miral/confine_pointer |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Brandon Schaefer (community) | Approve on 2016-09-23 | ||
| Daniel d'Andrada (community) | Approve on 2016-09-23 | ||
| Alan Griffiths | Pending | ||
|
Review via email:
|
|||
This proposal supersedes a proposal from 2016-09-21.
Commit Message
Add pointer confinement
| Daniel d'Andrada (dandrader) wrote : | # |
On 22/09/2016 05:26, Alan Griffiths wrote:
> Review: Needs Fixing
>
> Does not compile against Mir-0.20 - which is in 16.04 LTS, it needs similar compatibility hacks to MirPlacementGra
>
> /home/alan/
> auto confine_pointer() const -> mir::optional_
> ^
>
>
MIR_CLIENT_VERSION hasn't changed between mir 0.23 and 0.24 (where that
symbol got released). What do you suggest?
| Alan Griffiths (alan-griffiths) wrote : | # |
> MIR_CLIENT_VERSION hasn't changed between mir 0.23 and 0.24 (where that
> symbol got released). What do you suggest?
That's a Mir bug (not sure if I actually reported it).
I suggest detecting MIRCLIENT_VERSION (set by pkg_check_
if (MIRTEST_VERSION VERSION_LESS 0.24)
add_
endif()
| Alan Griffiths (alan-griffiths) wrote : | # |
> > MIR_CLIENT_VERSION hasn't changed between mir 0.23 and 0.24 (where that
> > symbol got released). What do you suggest?
>
> That's a Mir bug (not sure if I actually reported it).
>
> I suggest detecting MIRCLIENT_VERSION (set by pkg_check_
> CMakeLists.txt with:
>
> if (MIRTEST_VERSION VERSION_LESS 0.24)
MIRCLIENT_VERSION
> add_definitions
> endif()
| Daniel d'Andrada (dandrader) wrote : | # |
On 22/09/2016 05:26, Alan Griffiths wrote:
> Review: Needs Fixing
>
> Does not compile against Mir-0.20 - which is in 16.04 LTS, it needs similar compatibility hacks to MirPlacementGra
>
> /home/alan/
> auto confine_pointer() const -> mir::optional_
> ^
>
>
Fixed.
| Alan Griffiths (alan-griffiths) wrote : | # |
I'd like it fixed in a similar way to the window_
That way the downstream never needs to care about the Mir version - they just write against the MirAL API.
- 362. By Alan Griffiths on 2016-09-23
-
Fix build on yakkety
| Alan Griffiths (alan-griffiths) wrote : | # |
Hmm, debuild doesn't seem to like my libmiral1.symbols file. I wish I understood it.
- 363. By Alan Griffiths on 2016-09-23
-
Add pointer confinement to WM trace

Does not compile against Mir-0.20 - which is in 16.04 LTS, it needs similar compatibility hacks to MirPlacementGra vity. (A PITA I know, but we should support the LTS.)
/home/alan/ display_ server/ miral/include/ miral/window_ specification. h:196:57: error: ‘MirPointerConf inementState’ was not declared in this scope value<MirPointe rConfinementSta te> const&;
^
auto confine_pointer() const -> mir::optional_