ARM trust zone and Wayland on Genode?

[Norman Feske]

Hi Peter,

>    At the risk of coming across as a bit confused, For the life of me
> with Genode Labs having recognized the significance of actively taking
> into account the GUI layer in it's security model..    How in any real
> sense can Genode's growth road-map not yet openly account for Wayland's
> rise within the Linux computing world?....

the road map has a horizon of one year. It contains tangible goals to
which we commit ourselves. Most of the items you find on it are either
fundamentally important for us developers to use Genode as a
general-purpose OS (e.g., wireless, block cache), or are closely related
to our current research interests (in particular the work on base-hw).

We are nowhere near to catering the general-purpose computing needs for
a large crowd of normal users - such as people with a background of
being Linux power users. Advertising Genode as such would be a delusion.

Once we have reached the point where we developers are able to use
Genode every day, Wayland will possibly become very interesting to look
at. But today, it would just distract us from addressing more
fundamental topics.

As another technicality, it is almost futile to plan the integration of
Wayland before we even have a decent performing OpenGL implementation
(including GPU support) available. Enabling this is hard work though.
Maybe you'd like to lend a helping hand with this topic. ;-)

>   Am I just asking about
> something that Genode labs is choosing to play closer to there
> chest?....    I gather the good parts of Nitpicker might likely be a
> good starting point.  I'm just asking for some sense of assurance that,
> A. people are actively planing Genode's longer term growth future. B.
> that somehow we might be able to figure out how Genode might support
> apps that might be expecting a Wayland display compositor to talk to...

Concerning A, "long term" seems to be a pretty loose term. Our road map
covers just one year, which is a time frame that we can realistically
foresee. Personally, I wouldn't call it "long term".

I do not agree with your statement B as I am not aware of even a single
popular application that has a hard dependency on Wayland. Applications
are developed against toolkits such as Qt or Gtk. Since Genode supports
Qt, such applications can run natively on Genode already without the
need for Wayland. I don't expect this situation to suddenly change
within the next year.

As a general remark, if the use of existing software on top of Genode is
a concern, why not run Linux in a virtual machine on top of Genode as a
stop-gap solution? On NOVA, there exists a high-performance VMM called
Seoul. Additionally, VirtualBox will be available by the end of the
month. Those solutions can bridge the gap between the applications
Genode supports natively and commodity applications.

In short, Wayland is an interesting topic for sure but I don't think
that Genode will miss the train if we don't jump on the Wayland
bandwagon right now.

Cheers
Norman

-- 
Dr.-Ing. Norman Feske
Genode Labs

http://www.genode-labs.com · http://genode.org

Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden
Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth