RFC: plasma-framework IndicatorArea formFactor()

[David Edmundson]

On Mon, Sep 19, 2016 at 6:09 PM, Marco Martin <notmart at gmail.com> wrote:

> Hi all,
> I was looking at some systray-related bugs, such as
> https://bugs.kde.org/show_bug.cgi?id=365569
> and
> https://bugs.kde.org/show_bug.cgi?id=365967
> and it occurs to me they always have in common the same reason: even
> if I tried hard in the rewrite to make it less a "special case", it
> still is just that: a "special case" (ever having allowed plasmoids in
> the systray a long time ago in 2009 is still on of the design
> decisions i mostly regret but that's another story)
>
> what I'm thinking of, is having a new possible formFactor() value,
> IndicatorArea (why exactly FormFactor?because it's exactly that, while
> Horizontal means "can grow horizontally, constrained vertically, that
> would mean constrained both horizontally and vertically)
>

https://bugs.kde.org/show_bug.cgi?id=365967

It's a bug in that weather applet that it's rendering outside the area it
actually has - and that needs fixing anyway.
Introducing a new form factor won't solve that - in both cases you have to
do the exact same amount of work to that applet.

>
> So in that case applets would need to make sure to be square in that
> particular formfactor, like the weather that wants to eventually
> become non square in panels (and no, i don't want to allow non square
> icons in the systray.. ever)
>

> another difference would be (and that is kindof an hack, but at least
> it puts that immutability hack there and only there) containments of
> IndicatorArea formfactor would have a Mutable immutability°, because
> the systray itself should always be able to both add and remove
> plasmoids in itself even when immutable (even Systemimmutable) it
> could still hide/disable the ui to do so when immutable, but must be
> programmatically able to do so.
>

365618 and 365569 <https://bugs.kde.org/show_bug.cgi?id=365569> are
basically the same thing right?

You're saying we need mutability to be different in system tray containment
compared to it's parent. Sounds fine. +1

But I don't see why we would tie that to the formFactor.
Semantically they're not remotely related, the only link is that the system
tray happens to need both.


>
> Unfortunately as it needs changes in plasma-framework it would need to
> be a post 5.8 thing (how to properly fix 365618 in 5.8 I have no idea,
> except perhaps the systray applet itself trying to temporarly unlock
> plasma when it has to remove widgets, that would still be broken in
> systemimmutable case, but better than nothing)


> opinions? comments?
>
>
I don't object to the idea in principle, but the reasons you've listed here
don't support it.

David


> --
> Marco Martin
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20160919/e79ab64f/attachment.html>