other_stuff.h | [was: Review Request for KDecoration]

Martin Gräßlin mgraesslin at kde.org
Thu Nov 6 14:20:25 GMT 2014

On Thursday 06 November 2014 15:14:52 Thomas Lübking wrote:
> decoratedclient.h
> ------------------
> "maximal available" -> "maximal possible"
> "borderingScreenEdges" -> "adjacentScreenEdges" or "touchedScreenEdges"
> additional request:
> void monitorProperty(const QString &prop, bool onoff);
> signals: void propertyChanged(const QString &prop, const QVariant &data);
> (this allows further client/deco communication w/o having to know the
> backend or -eeewww- hook into the native eventfilter of KWin)

property means X11 property? In that case: good suggestion and clearly useful.

> decorationdefines.h, decorationsettings.h
> -----------------------------------------
> "BorderSize" - do we *really* want to keep this or rather allow
> pixel/pointwise configuration of 2,3, or 4 border sizes globally?
> decorationbutton.h
> ------------------
> nitpick: update() is overloaded explicitly while it's implicit in Decoration
> decorationshadow.h
> ------------------
> QSize topLeft() etc. etc.:
> a) shadow->topLeft() does not sound like a size
> b) the entire information (9 tiles) can also be provided by two rects
> (innerRect, outerRect) c) with a conv. func.
> sectionRect(Qt::WindowFrameSection), but I think those geometries actually
> only need to be calculated by the core? d) providing inner and outer rect
> may be "more safe", assuming that overlapping shadow sections are not
> desired? (atm. I cannot think of a variant that would cause anything but
> artifacts and we can easily check whether inner is really inside outer when
> setting them, evtl. combined?)
> Cheers,
> Thomas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20141106/30ab0c1d/attachment.sig>

More information about the kde-core-devel mailing list