Re: [rock-sm] 2005092818532619335 (ROCKCFG_TARGET_CRYSTAL_BUILDADDONS usage)
Stefan Fiedler <[email protected]>
| Newsgroups | gmane.linux.distributions.rock.devel |
|---|---|
| Message-ID | <[email protected]> |
Am Montag 14 November 2005 12:27 schrieb [email protected]: > Open Patch: 2005092818532619335 [by stf] + blindcoder > + stf > Stefan Fiedler: > remove target/crystal/ > postconfig.in because the code is already > in target/ > crystal.config.in and it prevents custom package selection > > VOTECHECK: Needs a positive vote from (fake), (clifford) or > (stefanp). > > (stf) of course it should read "... in target/crystal/ > config.in ..." > (dadchick) dsgsfd > (clifford) this code fragment is needed so e.g. the > linux26-header is built in the right stages. > > --------------------------------------------------------------------------- >- https://www.rocklinux.net/submaster/smadm.cgi?i=2005092818532619335 Hi Clifford, I don't get the connection to linux26-header. The code in target/crystal/postconfig.in, like the one in target/crystal/config.in, does the following: If ROCKCFG_TARGET_CRYSTAL_BUILDADDONS is unset, all non-CORE packages are disabled. If set, CORE packages will not be built in stage 9, and non-CORE package will be built only in stage 9. There's also a problem: this way it is not possible to build a CORE package against a non-CORE package (e.g. mplayer with libcaca output) in Crystal distributions. I'm also not sure about the purpose of ROCKCFG_TARGET_CRYSTAL_BUILDADDONS: besides the effects described above it excludes packages built only in stage 9 from the ISO image if set. Currently I think it's best to remove ROCKCFG_TARGET_CRYSTAL_BUILDADDONS, and exclude non-CORE packages from Crystal ISOs. With best regards, Stefan