GPL code in Unixlib

Peter Naulls peter at chocky.org
Sun Dec 26 02:03:56 PST 2004


In message <64ba46234d.Jo at hobbes.bass-software.com>
          John Tytgat <John.Tytgat at aaug.net> wrote:

> In message <3e27fb224d.peter at chocky.org>
>           Peter Naulls <peter at chocky.org> wrote:
> 
> > Unixlib currently contains a small amount of GPL code.  I'm keen to
> > replace this with LGPL or BSD style code.
> 
> Did we receive any particular requests for this ? Not wanting to start
> an endless licence discussion, but if we do the necessary actions to
> go for one licence, what are the realistic extra opportunities for the
> RISC OS community here ?

No, we haven't.  But given the small amount of code, it is a more than
reasonable target to achieve.  I also don't know what the extra
opportunities might be - but it's possible that having GPL code in UL
has been a sufficient barrier against people using what has now become a
very powerful C library for RISC OS.  Going for just once licence is
unlikely to be realistic - for example, glibc is subject to about 5 of
them.

> > _swi.s: Our _swi() implementation is taken from Straylight which is GPL.
> > I don't have any immediate suggestions on how to replace this.
> 
> A cleanroom implementation ? Or ask permission of Straylight to go for
> LGPL instead ?

I may be able to track down Mark Wooding, although I don't know how keen
he will be.

> Your observation is correct.  It is only a very small part of the _longlong.s
> code which came from the HAL sources.  Doing a quick look I think it
> only concerned the routines _ll_ushift_r and _ll_udiv10.  The latter
> is nothing more than using the same algorithm of _kernel_udiv10 and
> __rt_udiv10 but on 64 bit instead of 32 bit integers.

Ok, this sounds pretty easy.  Would you like to take care of this?  If
the correct response is to simply remove the GPL licence from the file,
than that is great.

> What about all the code contributed by the people mentioned in the
> gccsdk/unixlib/Docs/ReadMe3* files ? The code contributions of H. Rogers
> and Peter Burwood are substantional so should we contact them to clarify
> the status of their contributions ?

Oh, you mean "substantial".  Yes; although I think Nick can comment
further on this. As for the content of the files themselves, some of
these certainly need to be removed from being distributed - the
information they contain is desperately out of date.  

-- 
Peter Naulls - peter at chocky.org        | http://www.chocky.org/
----------------------------------------------------------------------------
Please Reply Properly - http://www.i-hate-computers.demon.co.uk/quote.html



More information about the gcc mailing list