Thanks Markus. To me, it would seem the image, with it's many classes,
is much more difficult to relicense than the VM.
See below

Markus Gaelli wrote:
> Andreas Raab told me after my quest to change the license "at least"
> for the current vms:
>
> - The VM is a binary and not source, thus the OSS discussion would not
> apply.
The VM is shipped binary, but the source is readily available. The
language seems ambiguous about the requirement to make VM source
available; even after modification of the VM (because, the modified
software would naturally be made available (in binary form) and satisfy
the license requirement)

> - None of the VM makers/maintainers is interested in this task at the
> moment, as the task is not important enough for them _now_
> - Changing the license of the VM might actually be _harder_ than
> relicensing the image code as all the plugins had to be looked through
I could see that. But, there are not a lot of plugins so finding each
owner and asking them if they are willing to license it openly (with
APSL2, or whatever the Squeak Foundation deems appropriate) should be
pretty easy. I remember reading Tim saying about his code, to the
effect: "sure"

But, what is the difference between "relicensing" and "changing the
license"? Seems the same to me.
> - If you wanted to help, you needed to classify all the changes to the
> VM from 1.1: Who did what and when... then you had to ask the authors
> for their agreement with a relicensing.
I could see that.
But, I wonder if we could: 
 * find out who did what as much as we can.
 * contact them for the license change.
 * Where we can't identify the code author, we advertise looking for the
author.
   Give it 1 month for them to reply. No reply means that code portion
is now open source.
* For those who refuse to relicense, the code is removed

If we don't do it this way, it may never get done.
>
> In the long future it looks like Ian Piumarta's new Pepsi and Coke
> Meta-VM might lift off:
Ian: How is your code licensed? Is it owned by VP?