On Thu, Apr 10, 2014 at 2:29 PM, Simon Kornblith <si...@simonster.com> wrote: > I'm not a lawyer either, but I don't think this is any more problematic than > the current situation with inclusion of Rmath etc. The combined work is GPL > (or LGPL, once we get rid of the GPL parts), but the vast majority of the > source files are (also) MIT. This seems fine since MIT gives all of the > permissions of (L)GPL and then some, and just because an author licenses > something as (L)GPL doesn't mean that author can't also make that file > available under another license. >
Right, but this needs to be spelled out in LICENSE.md. -Mike > On Thursday, April 10, 2014 1:31:24 PM UTC-4, Mike Nolta wrote: >> >> On Thu, Apr 10, 2014 at 11:07 AM, Tony Kelman <to...@kelman.net> wrote: >> > MPFR and GMP are LGPL, which is not quite as problematic or viral. >> > >> > Some of SuiteSparse is GPL, parts of it are LGPL, and at least one file >> > of >> > the Julia code in base for sparse matrices that is based on parts of >> > SuiteSparse is also LGPL. >> > >> >> IANAL, but doesn't this mean we've crossed the line from "work that >> uses the library" to "derivative work", and thus are in violation of >> the LGPL? Specifically, this part of section (2): >> >> "These requirements apply to the modified work as a whole. If >> identifiable sections of that work are not derived from the Library, >> and can be reasonably considered independent and separate works in >> themselves, then this License, and its terms, do not apply to those >> sections when you distribute them as separate works. But when you >> distribute the same sections as part of a whole which is a work based >> on the Library, the distribution of the whole must be on the terms of >> this License, whose permissions for other licensees extend to the >> entire whole, and thus to each and every part regardless of who wrote >> it." >> >> -Mike >> >> > >> > On Thursday, April 10, 2014 7:18:45 AM UTC-7, Jake Bolewski wrote: >> >> >> >> Are there any more liberally licensed libraries that get close to the >> >> functionality in MPFR, I know that there are some non-GPL BigInt >> >> implementations out there although I don't think any match GMP's >> >> performance. >> >> >> >> On Thursday, April 10, 2014 10:13:20 AM UTC-4, Stefan Karpinski wrote: >> >>> >> >>> There's been a lot of talk about turning various bits of functionality >> >>> like GMP, MPFR, FFTW and such into packages that simply happen to be >> >>> pre-loaded by default, making it easy to get a much more spare basic >> >>> Julia >> >>> version. This will definitely happen over the summer. Note that >> >>> OpenBLAS is >> >>> BSD-license, so using MKL is not necessary for non-GPL Julia. >> >>> >> >>> >> >>> On Thu, Apr 10, 2014 at 10:07 AM, Isaiah Norton <isaiah...@gmail.com> >> >>> wrote: >> >>>> >> >>>> Rmath too, but I think that is on the way out. All of the licenses >> >>>> are >> >>>> linked here: >> >>>> https://github.com/JuliaLang/julia/blob/master/LICENSE.md >> >>>> >> >>>> >> >>>> On Thu, Apr 10, 2014 at 10:05 AM, Jake Bolewski <jakebo...@gmail.com> >> >>>> wrote: >> >>>>> >> >>>>> As readline is now removed I think GMP (BigInt's) and MPFR >> >>>>> (arbitrary >> >>>>> precision floating points) are the only GNU ibraries left if you are >> >>>>> able to >> >>>>> use MKL and don't require FFTW. Steven also working on a branch >> >>>>> where he >> >>>>> provides FFT support in pure julia. >> >>>>> >> >>>>> Best, >> >>>>> Jake >> >>>>> >> >>>>> >> >>>>> On Thursday, April 10, 2014 9:26:55 AM UTC-4, Jay Kickliter wrote: >> >>>>>> >> >>>>>> There are bits and pieces in Github issues and posts, but can post >> >>>>>> a >> >>>>>> definitive list of what needs to be replaced/removed to make Julia >> >>>>>> non GPL? >> >>>>>> Will any functionality be missing? From what I understand I can use >> >>>>>> MKL for >> >>>>>> some stuff. I've read that MKL has the ability to mimic FFTW, but >> >>>>>> will Julia >> >>>>>> use that interface? >> >>>>>> >> >>>>>> For the record I'm not anti-GPL. I'd like to pitch Julia to my >> >>>>>> company >> >>>>>> as alternative to Matlab and C++. But our customers can't accept a >> >>>>>> project >> >>>>>> built with GPL. It's not a problem now, but I'm looking down the >> >>>>>> road when >> >>>>>> Julia can be compiled in to executables. >> >>>> >> >>>> >> >>> >> >
