Thanks guys, I'm going to add my two pence worth: I personally care about these things: 1) that I can acquire the source for pretty much everything on my computer, should I wish to; 2) that I can redistribute, modify etc any of these programs without seeking the owner's permission, or paying. 3) that I can have any/all of my own contributions recognised in this way, as can others.
That's for open source. Now for this particular library: 4) Windows support. It is *crucial* Windows support remains, because whilst most "supercomputer" clusters I know of run Linux or Unix, entry and even intermediate level work might well be done on a Mac or Windows. Similarly, young developers will almost certainly be learning on Windows - it's what universities teach on and what most people my age will be writing code for, initially. The same goes for older "new-entrant" programmers and anyone whose pc is "the family's pc". 5) Development community. I know what valgrind is, how to use gdb. I've assembled (gnu AS and NASM) gcd algorithms and I've read the x86-64 unix ABI. That said, this project is almost definitely going to be a steep learning curve for me - at my age (arguably any age) there is always more to learn and any other attitude is nonsense. From all the interactions I've had with the MPIR team, I see this as a community that will accept new developers and bring any and all contributions on board with nothing but positivity and enthusiasm. So for me, the choice of license isn't much of a biggie - provided it remains "open source" I'm not sure I care whether it's BSD, ASL, MPL, GPL, LGPL or whatever. What matters to me is the direction this project is going, which matches exactly what I need. So by all means upgrade to LGPLv3. That said, I don't think we should be "forced" into upgrading to LGPLv3 - we should do it for the right reasons. The issues with GMP are non-issues. The statement on the GMP website does not portray the full facts of that particular commit, for starters. The fact is, I could fork MPIR right now at LGPLv2 and any changes to MPIR's license would be rendered ineffective anyway. As could anyone, at any time. I think there is a more fundamental issue that GMP just didn't like their project being forked and supporting Windows as a platform, which isn't in the spirit of open source. Unfortunately for them, I could similarly fork their project at LGPLv3 and introduce Windows support and another development community too, as could anyone else. So, as far as I'm concerned, license isn't a big issue. Cross-platform/architecture support and development community is. That's what makes this project. Antony On 04/01/2010 08:41 AM, MPIR Team wrote: > Dear Developers, > > As many will be aware, we posted recently about the baseless and > unsubstantiated allegations on the GMP website and promised to > respond. > > We had anticipated that merely drawing attention to this material > would lead to unprecedented public outrage and negativity towards > MPIR. Of course, no publicity is bad publicity and we felt that this > latest advertising would be the best yet. But sadly, the wave of extra > developer effort we anticipated from this just hasn't materialised. > > After spending a few thousand dollars of precious funding on focus > groups and > taking prospective developers to dinner, we believe we've identified > the reason. The wave of hype did lead to a flurry of activity, but > just not in the anticipated direction. > > As a result of this research, we've made some significant changes to > our > strategy. In particular, we wish to make an announcement: > > In a desire to allow people to spend less time fruitlessly searching > MPIR releases for non-existent license infringements and hence more > time on new code for our benefit, we have decided today to consider > publishing future versions of MPIR under LGPL v3+ terms. > > We'd like the community to comment on this proposal. An announcement > of our first v3+ version of MPIR will be forthcoming in a few days. > > Best Wishes > > The MPIR Team. > > MPIR: "building our eMPIRe one bignum at a time" > "arithmetic without imitation" > > P.S: a <sponsor | http://tinyurl.com/y9dky9z> has expressed interest > in a different <license | http://tinyurl.com/yhlegke>. >
signature.asc
Description: OpenPGP digital signature