asymmetry?

[Brian Behlendorf]

On Wed, 13 Apr 2005, Danese Cooper wrote:
> On the subject of asymmetry...There is an assymetry in the warranty of
> provenance of many existing F/OSS licenses.  Original contributor is
> typically not required to make any sort of representation about their
> ownership of contributions nor about IP embedded in their
> contributions, while subsequent contributors often *are* required to
> make such representations.

I don't follow this.  Any "original contributor" most definitely makes a 
representation of the ownership of the IP by placing a copyright notice on 
the work.  If anything, many OSS projects are weak in doing the latter - 
getting assignments or license agreements from contributors.  Many OSS 
projects just assume that if you post that patch to the developer list you 
wanted it in the project.  But how many OSS projects are initially 
released without a copyright notice?

> In a project such as Apache where everyone starts on an even footing and 
> the seed code is derived from a 3rd source, this is a shared risk

Huh?

> but in a typical "corporate" project the asymmetry is concerning, 
> especially in cases where code might be encumbered (intentionally or 
> unintentionally) by large IP holders who might in future find offensive 
> enforcement actions to be an attractive strategy.

Wait - is the "large IP holder" the "corporate" who is releasing/managing 
the project (in which case, their license on the code isn't something that 
can be revoked later, at least I've never heard of a "we were just 
kidding" kind of attack) or are you referring to a different entity which 
might own patents that apply (nothing to be done about that case) or owns 
some of the code released by the "corporate" and never sanctioned its 
release (in which case the only guilty party is the "corporate" for 
improperly handling the IP of others)?

And what does this have to do with the concerns expressed here about 
licenses that establish an asymmetry in favor of the copyright holder? 
(put more broadly than Michael Bernstein defined it)

 	Brian