Jeremy Evans <[email protected]> writes: > Some other porters have complained that we have too many versions of > ruby in the tree. We currently have 7 ruby interpreters in the tree: > jruby, rubinius, and ruby 1.8, 1.9, 2.0, 2.1, and 2.2. So here is my > proposal for dealing with that. > > First, remove lang/rubinius, as it is slow because it doesn't work with > the in-tree LLVM, it is rarely used, and I couldn't get a recent version > to build with any of the in-tree compilers.
rubinius is now broken, following the SSLv3 removal. ok jca@ to delete it whenever you want. > This could be brought back > in the future if someone cares to do the work, but there needs to be a > plan so that it will always have a working LLVM. > > Ruby 1.8 will be sticking around for the foreseeable future as other > software depends on it, including amarok, subversion, and vim. > > For other ruby versions, change the OpenBSD support policy. Ruby > versions that will still be supported by upstream when -current is > released (currently 2.1 and 2.2 as 2.0 goes out of support in February) > will be fully supported by OpenBSD, and will have C extensions built by > default. When a ruby version will no longer be supported by upstream, > OpenBSD will stop building C extensions for the version, but it will > remain in the ports tree until the following version is no longer > supported by upstream, at which time it will be removed. > > This change in OpenBSD support policy would have the following effects > currently: > > 1) Remove lang/ruby/1.9 > 2) Stop building C extensions for lang/ruby/2.0. > > Thoughts on this? OKs? Your plan makes sense to me, but I don't know the ruby ecosystem well enough to have a meaningful opinion on this subject. :) > Thanks, > Jeremy > -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
