[TUHS] Caldera ancient UNIX license question

[imp@bsdimp.com (Warner Losh)]

On Aug 13, 2012, at 2:28 AM, arnold at skeeve.com wrote:

> Larry McVoy <lm at bitmover.com> wrote:
> 
>>> I'm
>>> sure there are other (mainly smaller) examples, though since we
>>> used no source-code control mechanism, tracing the details is
>>> non-trivial.
>> 
>> No SCCS?  When did Rochkind do SCCS?  Wasn't it early 70's?  I gotta believe
>> there is SCCS history out there.  And for the record, BitKeeper can read it.
> 
> I think Norman's point was that the Research guys didn't use a source
> code control system.  SCCS was around and documented in System III in 1980,
> so it was probably done before then, but not in the research group.

And even with a source code control system, it can be hard to know if there's an IP issue from commit logs, since they often are of the form "more" or "better" or "latest version" when there's isn't a culture of good commit messages.  And even when there is, if there isn't a good culture of documenting upstream sources, it can be hard.  And until at least a decade into the open source revolution there wasn't a general practice in the open source community about documenting upstream sources.

I defy you, for example, to identify with enough certainty to convince a corporate lawyer who actually wrote any of the code in Linux that's still around from the 0.9x or 1.0 time frame...  You can find all the tar.gz files from the time frame, and use tools to track which lines are still around, but knowing who actually wrote it can be tricky unless the patches hit a mailing list or had some other paper trail...  And the Linux development community was a lot more open and public than developments that happened 25 years ago to some IP that's changed hands a bunch of times...

Warner