[OE-core] The Mythical Sato Replacement

[Burton, Ross]

On 10 July 2012 09:56, Jack Mitchell <ml at communistcode.co.uk> wrote:
>> Did you consider QT instead of GTK+ ? I think having wayland would be
>> cool.
>
> I think Wayland support would be cool also. I think it would be a fantastic
> way to get people involved and encourage adoption.

Regarding Wayland, I covered that later in the email.  I really want
to see Wayland in oe-core, and the Shuku plan includes a Wayland
route.  I had a good chat with krh last night, the hardest bit will be
updating GTK+ to work with the latest Wayland.

Regarding Qt/EFL/wxWidgets/whatever, as Richard explained the
underlying principle is the most return on investment.  There are
several factors:

- the toolkit must be known by the person doing the work
- the toolkit must have a regular release cycle
- the toolkit must have ABI/API stability so that upgrading the
toolkit doesn't imply updating the UI
- working on an existing UI is better than starting from scratch

I'm willing to step up and do the initial work, and my preferred
toolkit is GTK+.  The Sato shell is sufficient for our requirements,
so the Shuku proposal is a low-investment plan that should give us a
refreshed UI for a few more years.

Sato has been on life-support for several years now, so if someone had
a pressing urge to write a replacement in Qt or EFL they've had plenty
of time to do it... and nobody has.

Ross