"Hauke Fath" <hauke%netbsd.org@localhost> writes: > Module Name: pkgsrc > Committed By: hauke > Date: Tue Jan 27 15:21:52 UTC 2015 > > Modified Files: > pkgsrc/net/coda: Makefile PLIST distinfo > Added Files: > pkgsrc/net/coda/patches: patch-coda-src_kerndep_coda.h > patch-coda-src_kerndep_pioctl.h patch-coda-src_librepair_repio.cc > patch-coda-src_resolution_rescoord.cc > patch-coda-src_venus_Makefile.in patch-coda-src_venus_fso_dir.cc > patch-coda-src_venus_venus.cc patch-coda-src_venus_venusrecov.cc > patch-coda-src_venus_vproc_vfscalls.cc > patch-coda-src_venus_worker.cc patch-coda-src_vice_Makefile.in > patch-coda-src_vice_codaproc2.cc > patch-coda-src_volutil_vol-ancient.cc > patch-coda-src_volutil_vol-dump.cc > Removed Files: > pkgsrc/net/coda/patches: patch-ac > TODO: Test on Solaris, SMF config files for venus and server TODO+= send all patches upstream add upstream bugtracker URL to all patch files Do you really have this working on any kind of solaris? Is there kernel support? Last time this came up I asked about that, and it seems it doesn't actually work. If the package doesn't actually work (fully, with kernel support), then I'd (as MAINTAINER) rather not have solaris patches, because they add update complexity and won't help. Arguably there should be ONLY_FOR_PLATFORM limited to NetBSD and Linux; I'm not aware of anything else working. If you can make it actually work, that's great and I'm glad to see it. If there is ever another coda release, and these patches cause grief, I'm just going to nuke them if they haven't been filed upstream. The point here is that pacakges which need help should be fixed upstream, rather than in pkgsrc. (If upstream doesn't deal, fine, but I think it's required to try, and that's I believe documented in The Guide.) Also, it seems like there are a lot of things rolled into one commit; I'd prefer to see separate commits for separate logical changes (like the CONF_FILES stuff). (I would have said this if asked.)
Attachment:
pgpeSLdv8rEgG.pgp
Description: PGP signature