Review Request: reasonable applets size in bigger (probably vertical) panels

[Marco Martin]

On Monday 04 February 2008, Aaron J. Seigo wrote:
> On Sunday 03 February 2008, Marco Martin wrote:
> > -----------------------------------------------------------
> > This is an automatically generated e-mail. To reply, visit:
> > http://matt.rogers.name/r/49/
> > -----------------------------------------------------------
> >
> > Review request for Plasma.
>
> unfortunately this has already been submitted so i can't coment on review
> request anymore, but this patch is incorrect in several interesting ways.
> =)

ok, first of all i'm sorry of messing up things
it seemed sooo nice, well :(

> first the obvious stuff:
>
> * this will not work in any applet which reimplements contraintsUpdated
> since Applet::constraintsUpdated won't be called. instead, such code should
> go into flushUpdatedConstraints (there's already some stuff like there in
> there) i'll add a comment in Applet::constraintsUpdated so that this
> doesn't get repeated in the future, as i can see how it's easy to end up
> with such a bug as it's not immediately obvious.
woops, didn't know

> now the non-obvious stuff that we probably need to discuss:
>
> * i now have tiny little icons for, e.g. the device notifier. ditto for the
> kickoff button. but for other icons. this new inconsistency, dependant on
> d->square, is not great.
>
> * the code is not run if d->square, but d->square impacts
> expandingDirections() so isn't necessary as a check. an update of the
> widget *is* necessary however.
>
> these two items can be resolved by simply removing the check for d->square.
> that, however, leads to a new world of hurt.

so for appearing right the square applets must be and square and centered, so 
something like in a layout that centers them... aaaah my head explodes :)

> * why rely on the size of an icon for this? what about applets which aren't
> icons and have no relation to them? putting the code in
> flushUpdatedConstraints higlights this problem very nicely with, e.g., the
> pager which is now cut off and the battery applet which is now squished up.
>
> * the resulting icons in Horizontal layouts is overly tiny.
so making this check only to vertical layouts? not good

or putting a different maximum size bigger by some rule? like on how much the 
screen it's wide than tall? (yeah, i know sucks on multi monitors)
>
> so ... looking at the problem some more ... it seems that the real issue is
> two fold:
>
> * it only applies to certain classes of applets which should conserve
> space. * it is commonly related to the size of icons, but probably to any
> item which behaves similarly (e.g. a fixed representation that simply grows
> to fill the space)
that's why i had tought to applets with no expanding direction, but really i 
can't think on a solution for square applets, if a battery would end up 
stretched the applet would be much more complex to keep the battery svg 
square, there it is always thet idea of the horizontal layout that keeps 
insisting in my mind but i know it sucks...

Cheers,
Marco Martin