[TUHS] Source code abundance?

[Warner Losh]

On Wed, Mar 1, 2017 at 1:28 PM, Clem Cole <clemc at ccc.com> wrote:
> Below
>
> On Wed, Mar 1, 2017 at 2:29 PM, Steve Nickolas <usotsuki at buric.co> wrote:
>>
>> On Wed, 1 Mar 2017, Henry Bent wrote:
>>
>>> My understanding is that System V source of any sort is not legal to
>>> distribute.  I believe that source exists and has been archived for at
>>> least some variants of SVR1, SVR2, SVR3, and SVR4.
>>
>>
>> Well, that's probably 95% true...the other 5% is Solaris. ;)
>
>
>
> Not a lawyer and don't play one TV or anywhere else....
>
>
> Some thoughts...
>
> 1.) UCB Regents Position per AT&T/BSDi/UCB - anything through 32V is public
> domain (see groklaw)
> 2.) All of Sun & IBM, bought out source licenses from AT&T with rights to do
> anything.... IBM is based on the SVR3 license, Sun on SVR4
> 3.) IBM's license is the basis for the OSF/1 license
> 4.) HP independently eventually gets is own bought out license, but I'm not
> sure what it's based [need to google the old UNIGRAM/X or the like]
> 5.) Sun takes SVR4 in and starts to add "Solaris features" to it (not going
> to argue percentages here for the moment).
> 6.) Sun open sources this code base...
>
> Now some questions....
>
> From the above, one could argue that set of code included in Solaris from
> the SysV linage was made public by step 6.
>
> I have seen argument that anything through SVR3 is public because of the
> actions of IBM, HP, and SUN when the code was bought out; but I have not
> seen a definitive action like step 6 that infer all of SVR3 was public.

I would be skeptical of that assertion. Copyright law doesn't allow
one to gain rights for earlier versions of a work they got rights for,
except to the extent that the earlier work is wholly included in the
later work.

Warner