Sublicensing

Daniel Vainsencher danielv at netvision.net.il
Sat Aug 16 23:41:40 UTC 2003


Peter Crowther <peter at crowther.demon.co.uk> wrote:
> > From: [...] Daniel Vainsencher
> > I like Colins answer - if the implementation is based on new ideas, it
> > obviously isn't copied, so no real risk.
> 
> - Will the *entire* implementation be based on new ideas, or will even one
> small part (say, one method) not be?  That small part *may* be sufficient to
> trigger a suit from an organisation that does not necessarily know what it
> is talking about.  We may have cut the risk of this by as much as a factor
> of ten - probably not more than that if it could be claimed that there is
> *any* derivative code in the 'clean' system.  Is this sufficient?  Or is the
> risk of any organisation dragging a developer through the courts still too
> high, as they have money and (s)he does not?
I don't think we're communicating here...
1. I am not interested in false suits - there is nothing we can do to
prevent those but stop existing.
2. I don't know what you mean by a method not being based on new ideas.
If you design something from raw ideas (not code) and write the code it,
the copyright is yours. To say that someone may dispute this returns us
to point 1.
3. Obviously we need to be pretty conservative about what we consider
"replaced". A new package might have to separated into two parts, one of
which is clean, and the other, while new, would need to be considered
old because the line isn't clear enough. We'll still make progress.

> - Will a piecemeal rewriting be sufficient, or will it all have to be done
> in a 'big-bang' fashion?  Remember the quote: 'He's had the same broom for
> thirty years.  It's had six new heads, five new handles, but it's the same
> broom."  A piecemeal rewrite of Squeak is [perceived as] the same Squeak,
> subject to the same attacks.
What basis do you have for saying this? If I publish a package on SM,
that doesn't contain code from the image, and this is merged into the
image, I don't lose copyright over my publication. I can reuse it in a
future cleanly licensed Squeak. This is what I propose we start doing.

Daniel



More information about the Squeak-dev mailing list