This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch, gfortran] pr22570 and related issues.


Jack Howarth wrote [on the NIST suite license]:
Please note that within the United States, copyright protection, under
> Section 105 of the United States Code, Title 17, is not available for
> any work of the United States Government and/or for any works created
> by United States Government employees. User acknowledges that these
> test suites contain work which was created by NIST employees and is
> therefore in the public domain and not subject to copyright. The User
> may use, distribute, or incorporate these test suites provided the
> User acknowledges this via an explicit acknowledgment of NIST-related
> contributions to the User's work. User also agrees to acknowledge, via
> an explicit acknowledgment, that any modifications or alterations have
> been made to these test suites before redistribution.

Not being an expert on GPL, those terms seem close enough to GPL that NIST could easily relicense the packages using those terms under GPL. Of course going those all the adminstrative hoops to do that would be a lot of work for sure.

That's a really weird definition of "public domain" that they seem to be using in those terms. If it's really in public domain, we should be free to do whatever we want to with it -- GPL it, copy it, sell it, whatever -- except for claiming our own copyright on it, and the requirement to acknowledge NIST's contributions is a moral one with no legal teeth whatsoever. (Insert usual "but I'm not a lawyer so I don't know for sure" disclaimer here.)


My guess is, if there's worth the bother of incorporating it into the gfortran test suite, we should send off an email to FSF's legal team asking them for a confirmation on the matter, but it certainly appears to my non-lawyer eye from that "license" that a "User" would be free to give the code (in, presumably, a modified-to-fit-the-testsuite form) to FSF with no restrictions whatsoever, and FSF could then quite legitimately distribute it with the rest of gfortran's test suite.

(I should also think that a comment in the code should be sufficient to fulfill the moral obligations of referencing NIST. And, if it's a moral obligation rather than a legal one, there's no requirement that it be passed along in succeeding license.)

- Brooks


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]