Implement support for xdg-activation #426
Conversation
Drakulix
force-pushed
the
feature/xdg-activation
branch
2 times, most recently
from
070cb9f to
f41bec4
Compare
src/activation.rs
Outdated
| /// *Warning*: Many compositors will issue invalid tokens for requests without | ||
| /// recent serials. There is no way to detect this from the client-side. | ||
| fn serial(&self) -> Option<(&wl_seat::WlSeat, u32)>; | ||
| /// Surface of the window requesting the token, if applicable. | ||
| /// | ||
| /// *Warning*: Many compositors will issue invalid tokens for requests from | ||
| /// unfocused surfaces. There is no way to detect this from the client-side. | ||
| fn surface(&self) -> Option<&wl_surface::WlSurface>; |
There was a problem hiding this comment.
No need to use Option for both of them, just enforce the use of what should be passed..
There was a problem hiding this comment.
Yes there is use in requesting a token without providing these.
There was a problem hiding this comment.
I mean, what could it be, it's all non-documented. You better use app_id if you want to have a special logic in compositor.
There was a problem hiding this comment.
E.g. requesting a token without an active surface. Which might be valid for privileged clients.
Because it is all non-documented, we should actually expose everything the protocol allows instead of assuming a certain compositor policy.
src/activation.rs
Outdated
| /// *Warning*: Many compositors will issue invalid tokens for requests without | ||
| /// recent serials. There is no way to detect this from the client-side. | ||
| fn serial(&self) -> Option<(&wl_seat::WlSeat, u32)>; |
There was a problem hiding this comment.
You're returning seat and serial from a function called serial, sounds really odd.
I'd suggest to separate seat and serial. Also &Wlseat should be better.
There was a problem hiding this comment.
That's the naming the protocol uses and it expects either both or none of the above. So I can't really make two methods out of it. We could just change the name.
There was a problem hiding this comment.
I mean, whatever protocol uses is up to the protocol, we're writing a toolkit which should have saner names.
There was a problem hiding this comment.
I guess... Still like separate serial and seat a bit more.
There was a problem hiding this comment.
Well, if you tell me what the implementation is supposed to do, if it gets either but not the other... just don't set it? That feels wrong, if we can enforce it on a type level.
Drakulix
force-pushed
the
feature/xdg-activation
branch
2 times, most recently
from
67adadd to
db1c2d3
Compare
src/activation.rs
Outdated
| /// *Warning*: Many compositors will issue invalid tokens for requests without | ||
| /// recent serials. There is no way to detect this from the client-side. | ||
| fn serial(&self) -> Option<(&wl_seat::WlSeat, u32)>; |
There was a problem hiding this comment.
I mean, whatever protocol uses is up to the protocol, we're writing a toolkit which should have saner names.
Drakulix
force-pushed
the
feature/xdg-activation
branch
3 times, most recently
from
5523cb2 to
ed0b5e6
Compare
Drakulix
force-pushed
the
feature/xdg-activation
branch
from
ed0b5e6 to
6a1fa55
Compare
Drakulix
force-pushed
the
feature/xdg-activation
branch
from
6a1fa55 to
0be6b8d
Compare
Implements xdg-activation.
The second commit adds it to the simple-window example as a test. Given how heavily commented that is, I guess that is not really a good place to put it. But without a proper window it doesn't make a whole lot of sense to use it. I could also just drop that commit?