Libiberty license roundup (questions/potential problems)

[Mike Stump]

On Friday, May 23, 2003, at 03:59 PM, Nathanael Nerode wrote:
> Looking at libiberty, it seems to be under a mass of different 
> licences.

Yup.

> There are several problems here.  The first are the 
> (non-autogenerated) files
> with no explicit copyright notice or licence.  I would assume that 
> they were
> under the terms of the "rest of libiberty", except that it's not clear
> what that is.

Someone would need to go back to the Cygnus devo tree and find out who 
and when they were checked in originally, and what files around that 
time the person was checking in and what status those files had.  Steve 
checked in some of the files, and wrote some, quite a bit of those were 
PD.  He also lifted some from BSD land.

In general, we should split the source internally into two, those that 
are BSD/PD/GPL with libgcc exception, and another directory with 
LGPL/GPL code.  I think that we should have two libraries, one for each 
of these directory hierarchies.

Clarifying the result of this research into explicit terms in the files 
I think would be good.

Clarifying that the files are part of libiberty I think would be good.

> * No license, University of California copyright
>   xatexit.c

I think this is supposed to be under a BSD style copyright.

> * No license, FSF copyright
>   vfprintf.c

This was written by us for us.  I think it should be GPL with exception.


Another way to clean it up, would be to check FreeBSD/NetBSD for 
corresponding versions of the routines, and replace the one in 
libiberty.