[ros-kernel] Code Copyright Issues

[Vizzini]

By the way, in case it wasn't obvious, I am not a lawyer; I'm just
trying to make sure our policies (right or wrong) are clear to all.

WRT what Andrew said (and I think there are many others who would say
the same thing he just said):

On Tue, 2004-02-10 at 14:54, Andrew "Silver Blade" Greenwood wrote:
> > Our code contribution policy requires 100% original typing by the
> > programmer.  You cannot use copy and paste for anything, 
> 
> I often find myself looking at MS sample code, and then typing it out again,
> in an almost identical structure (procedure-wise.) 

If you have to duplicate the ideas in code for technical reasons (which
would include "it is better this way because of x"), that's OK.  I would
discourage you from duplicating the code structure of MS's code for no
good reason, though.  

Interfaces are another story, of course.  Your functions have to have
the same names, the same calling conventions, the same return values,
the same arguments with the same names and types in the same orders,
etc.  These things are necessary for compatibility, and are just
expressions of the interface.  

Structures, global variables, typedefs, enumerations, etc.  These all
have to be named exactly the same way, have members with exactly the
same names and values, and so forth.  Remember, copyright covers the
expression of an idea, not the abstract idea itself.  

For some light reading, google about the Altair case, the Borland vs.
Lotus case, the MS Excel case, AMD vs. Intel, the Visicalc case, and
others.  Nothing gets directly to header copying, but these cases are in
the same ballpark.

> And for things like
> switch statements, I have found copying and pasting the case labels a
> time-saver. 

Yep, I'm sure it saves time.  I really think the policy has to be pretty
hard-lined, though.  The more in-between cases we allow for, the more
vague the legal standing of the code is.  Don't copy and paste, and then
it won't be a question.

<snippage about another driver>

I was not sufficiently clear about one thing:  if you would like to tell
me about code you know of that is in violation of copyrights, please do
so off-list, to me personally (or to Jason Filby, if you prefer).  That
sort of thing is not appropriate content for a mailing list.

In that vein, please *do* let me know (privately) of anything that you
know of that violates a copyright.  I will be developing a strategy for
scrubbing up our codebase in the not-so-distant future, and I'd like to
be armed with a list of trouble spots.

Thanks.

 -Vizzini