[PATCH weston-ivi-shell v5 9/9] Modify example clients to support ivi-application.xml

[Pekka Paalanen]

On Tue, 18 Mar 2014 23:57:32 +0900
Nobuhiko Tanibata <NOBUHIKO_TANIBATA at xddp.denso.co.jp> wrote:

> Signed-off-by: Nobuhiko Tanibata <NOBUHIKO_TANIBATA at xddp.denso.co.jp>
> ---
> 
> Changes for v2, v3 and v4
>   - nothing. Version number aligned to the first patch
> 
> Changes for v5
>   - support weston-dnd-ivi to verify wl_pointer::set_cursor and wl_data_device::start_drag
> 
>  clients/.gitignore   |  6 +++++
>  clients/Makefile.am  | 72 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>  clients/simple-egl.c | 70 ++++++++++++++++++++++++++++++++++++++++++--------
>  clients/simple-shm.c | 53 +++++++++++++++++++++++++++++++++-----
>  clients/window.c     | 45 ++++++++++++++++++++++++++++++--
>  5 files changed, 228 insertions(+), 18 deletions(-)

I don't like the #ifdeffery, but any other solution to achieve the same
would probably copy a lot of code. There is also the controversial
question of keeping the simple clients simple while still being useful.
If they don't support ivi-shell, they won't be useful on ivi-shell
environment.

But I suppose we have kind of crossed that line already.

What I wonder is, if it would be better to just add ivi-shell support
to these two simple clients and toytoolkit as a first class feature.
Not by #ifdeffing and compiling another set of executables, but by
having the support always built in, and detect the shell at runtime.


Thanks,
pq