x11-libs/openmotif-2.3.3-r1 adds a dependency on dev-util/byacc. I had to drop the following keywords: arm sh x86-fbsd ppc-aix ia64-hpux x86-interix x64-solaris x86-solaris Please re-keyword. Thanks in advance.
One thing to remember: The license of openmotif disallows its use on nonfree platforms.
(In reply to comment #1) > One thing to remember: > The license of openmotif disallows its use on nonfree platforms. Well, at least dev-util/byacc is free software and can be tested and keyworded. And as I understand our keywording policy, that's also enough for re-keywording of openmotif. (Imagine that byacc had already been keyworded before openmotif was bumped. In that case, openmotif's keywords wouldn't have been dropped, in the first place.)
~arm done
~x64-solaris ~x86-solaris done
Keyworded sh by armin76, thanks. Still missing: x86-fbsd ppc-aix ia64-hpux x86-interix
~x86-fbsd done. Added sys-freebsd/freebsd-ubin as alternative for dev-util/byacc.
(In reply to comment #0) > x11-libs/openmotif-2.3.3-r1 adds a dependency on dev-util/byacc. Is dev-util/byacc something different than virtual/yacc? > ppc-aix > ia64-hpux > x86-interix Done for dev-util/byacc. However, don't see fit for openmotif due to the license (use lesstif instead), maybe drop them from the old versions too?
(In reply to comment #7) > Is dev-util/byacc something different than virtual/yacc? Yes, it explicitly requires byacc instead of bison -y, because the bison-generated parsers cause problems. See for example bug 312719 and bug 355795. Feel free to debug why it doesn't work with bison. ;-) > > ppc-aix > > ia64-hpux > > x86-interix > > Done for dev-util/byacc. > > However, don't see fit for openmotif due to the license (use lesstif > instead), I don't think that we have a policy that would forbid keywording of proprietary packages. KEYWORDS just states that the package (or at least a previous of it) works on such and such architectures, i.e. this variable is of purely technical nature. For any license issues we have the LICENSE and RESTRICT variables. I believe that the ebuild properly assigns both of them. Especially, the package is fetch-restricted on systems for which the license doesn't allow free distribution. (See also bug 334261 for a longer discussion.) Therefore I've re-added the previously dropped keywords now.