Hello Tim,
* Tim Newsham <tim.newsham@...9...> [2014-12-22 07:06:04 -1000]:
my two cents, tinted by my own personal world view (I work in the security field) -- I think a great way to showcase Genode's strengths (and the strengths of underlying kernels, such as (soon) seL4) would be to port the chrome browser. It's open source (and unfortunately very large), it's one of the most important programs today, it is already designed for security, and one of its strengths is its sandboxing of components. However, chrome's sandbox is not nearly as strong as it could be, and a capability based system like Genode that can put together the minimal set of necessary components would make a much stronger sandbox. Chrome is already architected properly for this, it would just needed to be ported appropriately with the right platform support.
I agree, Chrome would be a great showcase, indeed. However, as you already mentioned, it is very large and quite complex — porting it would require much effort. Unfortunately, we currently do not have the time to perform this task. That being said, we are willing to support anyone who tries to port Chrome to Genode as best as we can.
At the moment it would be more feasible to send Arora into an early retirement and use a more modern or atleast actively maintained Qt5 based browser like Qupzilla instead. An older version of it was already ported to Genode by ksys labs and porting the latest stable version should not be that big of a deal.
Cheers Josef