Hi Michael,

On 3/14/07, Michael van der Gulik <mikevdg at gmail.com> wrote:
> Hi all, especially Pavel.
>
> What are your long-term plans for KernelImage?
>
> I'm working on a more secure version of squeak, and KernelImage looks like a
> good starting point.
>
> Pavel: would you want the following features in KernelImage, most of which
> would break backwards compatibility:
>
> - Hierarchical Namespaces,
>
> - Preventing multiple applications in an image from being able to affect
> each other (in terms of excessive memory usage etc),
>
> - Capability-based security, meaning that core Squeak classes cannot be
> modified, devices must be granted access to, etc, enabling you to load
> foreign, untrusted code and execute it securely.
>
> If so, then I'd like to work with you on this. If not, I'll have to fork
> KernelImage for my own project.

The KernelImage is not designed as a fork of Squeak. Contrariwise it
should help to share code between current Squeak forks. So if the
Squeak community will want and accept this changes then it will be
part of the KernelImage. Current KernelImage itself is only an
evolutional step to better Squeak modularization (~Spoon).

So if you really want this, you will have to fork - it would be nice
if it will be possible to have it as the optional functionality. Then
it will be easier to accept it in general.

> I assume you've released KernelImage under the MIT license?

My changes are under MIT license but the original code is still under Squeak-L.

Cheers,
-- Pavel


> Cheers,
> Michael.
>