Unanswered comment on the PurePrimer1 page
Jiri Spitz
There is also an unanswered comment on the PurePrimer page:
Have you considered using MPIR rather than GMP? It is binary compatible,
but the development is more active, with support for newer CPUs.
Have you considered adding MPFR and MPC libraries to permit
multi-precision floating point complex numbers?
Jiri
Roman Neuhauser
> There is also an unanswered comment on the PurePrimer page:
>
> Have you considered using MPIR rather than GMP? It is binary compatible,
> but the development is more active, with support for newer CPUs.
licensed" mean? http://www.mpir.org/#license says "MPIR is overall
licensed LGPLv3 or later at user's discretion". does that mean some
bits are GPLed? proprieatary? they don't say. i'm not trying to be
difficult, it's just that that "overall" has really made me curious.
--
roman
Albert Graef
> i checked the MPIR site and my impression is wtf does "overall
> licensed" mean? http://www.mpir.org/#license says "MPIR is overall
> licensed LGPLv3 or later at user's discretion".
Yeah, that makes me a bit wary, too. Since the bigint library is linked
into the runtime, we need something that's compatible with LGPLv3+,
period. Also, GMP is shipped by most distributions and has reached a
certain level of maturity.
I wonder what the motivation behind this fork is. I couldn't find a
clear statement about that on the website either. The only remark that
might give a clue: "Note, it is not necessary to assign copyright to the
Free Software Foundation in order to contribute to MPIR." Well, maybe
they had frustrating experiences getting their patches into GMP, and
wanted something that allows better community involvement. That might be
a legitimate reason.
If MPIR is really binary compatible as they claim (source compatibility
is already enough), we could certainly add a configure switch to support
it, that would be easy. But I guess I'll procrastinate on this a bit
more until a Pure user actually requests it (and preferably provides a
patch to configure.ac).
Albert
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.G...@t-online.de, a...@muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikinformatik.uni-mainz.de/ag
Albert Graef
> i checked the MPIR site and my impression is wtf does "overall
> licensed" mean?
Well, maybe it's some of the demo programs which are under the GPL, just
like in GMP. But they should really spell out their licensing terms on
the website so that they are 100% clear.
(They will have to use the same terms as GMP anyway, as a fork it's
probably still mostly the original GMP code.)
I renamed this thread so that we can further discuss MPIR or MPFR
support here, if anyone's interested in that.
jason
I'm one of the developers of MPIR and I can confirm that the MPIR
library is LGPL v3 or above. Originally we started out with MPIR being
LGPL v2.1 or above , however later on some of the code that we wanted
to include was only licensed for v3 or above , so the only way we
could include the code was to change the overall license of MPIR to v3
or above. It's possible in the future if we replace all those files
that we can go back to being v2.1 or above. In the latest version v2.2
( to be released next week) the demo programs have been removed ,
although if you dont use them (your'll know if you use them as they
have to be compiled separately) , you dont have to worry about their
license.
Jason
On Nov 10, 7:43 am, Albert Graef <Dr.Gr...@t-online.de> wrote:
> Roman Neuhauser wrote:
> > i checked the MPIR site and my impression is wtf does "overall
> > licensed" mean?
>
> Well, maybe it's some of the demo programs which are under the GPL, just
> like in GMP. But they should really spell out their licensing terms on
> the website so that they are 100% clear.
>
> (They will have to use the same terms as GMP anyway, as a fork it's
> probably still mostly the original GMP code.)
>
> I renamed this thread so that we can further discuss MPIR or MPFR
> support here, if anyone's interested in that.
>
> --
> Dr. Albert Gr"af
> Dept. of Music-Informatics, University of Mainz, Germany
> WWW: http://www.musikinformatik.uni-mainz.de/ag
Albert Graef
> I'm one of the developers of MPIR and I can confirm that the MPIR
> library is LGPL v3 or above.
Hi Jason, thanks for clarifying this. LGPLv3+ is fine.
> In the latest version v2.2
> ( to be released next week) the demo programs have been removed ,
> although if you dont use them (your'll know if you use them as they
> have to be compiled separately) , you dont have to worry about their
> license.
I agree. There's really no need to remove the demo programs, we were
just worried a bit about that "overall license" wording on your website.
If you just say that libmpir is LGPLv3+ then everything should be fine;
a note about GPL'ed demo programs can be in the fine print, or will be
obvious from the license notices at the beginnings of the programs anyway.
I've added a --with-mpir option to Pure's configure now, for those who
want to give it a try. Pure seems to work fine (passes all checks) when
using MPIR.
I'm not sure, though, whether linking the interpreter against libmpir
might cause hiccups with some of the addon modules which are linked
against libgmp. But, as mentioned on the MPIR website, you can always
just build MPIR with the --enable-gmpcompat configure option and use it
as a drop-in replacement for GMP. If you do this, the --with-mpir option
isn't needed when configuring Pure. I'm currently testing this setup and
it seems to work fine so far.
Albert
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany