>> * I would rather have an expanding property in favor of stretchFactor. How
>> often do we need to say that item A should expand a bit more than item B? (I
>> don't do that very often). Stretch factors adds also several possible
>> behaviors. (Which one is the best? Which one does the end user expect?)
>> (QBoxLayout and QGraphicsLinearLayout differs in this respect).
>
> Well, the idea is that expanding is handled by vertical/horizontal
> size policies, like:
> Item {
> Layout.horizontalSizePolicy: Layout.Expanding
> }
> The stretchFactor is just a bonus and it's not tied to the Item
> definition. Basically, it's related to a specific layout. For example:
Perhaps just the name Layout.sizePolicy is precise enough. Is there a case
where it is relevant to set both in the same layout as long as you still have
to specify Row or Column?
> I can remove the 'spacing' property from RowLayout and ColumnLayout.
> It was just a fast way to provide the fixed spacer use case.
I would suggest that we keep it. Partially I want to keep it more or less
compatible with row/column and I almost always want some default spacing
between widgets. In fact I think the default spacing should be set to 4 pixels
and not 0 by default. (unless we want per-component style based values) Apart
from that, I would suggest that stretching is off on items by default. (Keep
stretchFactor at 1.0 as default but ignore it unless Layout.sizePolicy ===
Layout.Expanding.)
I am currently a little bit puzzled by the behavior when no items have a
stretch factor. I wonder if we should insert an invisible expanding item
implicitly at the end?
The other thing that was odd was having to set Layout.implicitWidth on an Item
to create a spacer. For some reason I expected just setting width to work. But
I can see how that makes sense though… Perhaps a convenience Spacer{} would be
worth it.
> layouts, which was the first target, but that can be changed. I just
> wish to have more spare time to get on the Grid implementation :).
I am happy that you are managing to find some time to work on this though. It
is already very useful.
Regards,
Jens
_______________________________________________
Qt-qml mailing list
[email protected]
http://lists.qt.nokia.com/mailman/listinfo/qt-qml