D11250: Expose presentation via MPRIS to rich remote controllers

[Friedrich W. H. Kossebau]

kossebau added inline comments.

INLINE COMMENTS

> aacid wrote in dbusabstractadaptor.h:32
> This seems like random glue code, wouldn't this make more sense in Qt or in KF5?

Possibly. Similar code working around the missing Qt feature is found across code. Actually one would expect this indeed with Qt, but the one known related bug report https://bugreports.qt.io/browse/QTBUG-48008 got for whatever reason a harsh wontfix.
Not my  playing ground, so I escaped into manual workaround like found elsewhere. Will be happy to have something in Qt or KF5, but do not have resources for that.

> aacid wrote in mprismediaplayer2player.cpp:58
> No, you can not, file can change at any moment.

Also during presentation? Okay. Any change signal you would recommend to catch here? Would encode some thumbnail versioning then to ensure uniquness.

> aacid wrote in mprismediaplayer2player.cpp:304
> Why do you need that?

To avoid creating repeated requests for the same pixmap.

Challenge I see: I need to get a thumbnail per page. To speed-up things, it will be done only for pages currently visited. But given the request is processed async, when a user switches pages while the requests are still on-going, when entering the same page multiple times multiple requests would be created, one per visit.

Is there some internal mechanism to squash such requests?

> aacid wrote in mprismediaplayer2player.h:99
> are this ugly uppercase tyied to the spec? Isn't there like a magic way in Qt to create the api from the xml spec?

Yes, needed due to using `sessionBus.registerObject(mediaPlayer2ObjectPath(), this, QDBusConnection::ExportAdaptors);`, and D-Bus methods should be Uppercase and so does MPRIS.
 I only know `qt5_add_dbus_interface()` for creating nice-Qt-style API for interfaces to be used with remote objects. For local actual implementations of interfaces I am not aware how to avoid that, the other code I saw doing MPRIS implementations did things the same way as here. Might have missed something, but done to my best knowledge.

REPOSITORY
  R223 Okular

REVISION DETAIL
  https://phabricator.kde.org/D11250

To: kossebau, #okular
Cc: aacid, pino, rkflx, #kde_connect, michaelweghorn, ngraham
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/okular-devel/attachments/20180318/1afcaeb8/attachment.html>