The idea of sig files is to capture user intent/desire as opposed to an exact list of installed packages. If a "content" package will always be installed indirectly as a dependency of a GP then we need a GP sig file only for the GP. If a package is likely to be installed directly by the user then it should have a "loose package" sig file.
If you are saying that docs and g11n should be pulled in by ha-cluster-full rather than ha-cluster-framework-full, then I agree. $0.02 --emk Jonathan Mellors wrote: > Hi Ed, > > It's still open for discussion. In my reply to Nick, I wrote: > "I think the sub-group -full needs to contain docs and g11n, otherwise > there's no consistency when you look at quorum server and later on geo." > > We also need your input on whether the docs and l10n packages should > have signature files. > > Thanks > Jonathan > > > Ed McKnight wrote: >> Sorry so late. Did we agree whether man pages, etc, will go in >> ha-cluster-full vs ha-cluster0framework-full? No remarks beyond >> those already made. >> >> hth, --emk >> >> Jonathan Mellors wrote: >>> Hi Folks, >>> >>> Here's a couple of webrevs for the importing of the external SVR4 >>> packages from Docs and Localization into the IPS repository during >>> the build. This functionality is only required until Docs and L10N >>> can build their own IPS packages directly. >>> >>> This would have been a relatively small change, if it wasn't for the >>> fact that I had to move the "install: all ipssend" rule from >>> ipsdefs/Makefile.targ into the individual package Makefiles, so that >>> docs and l10n could use a different install rule. >>> >>> >>> Core: >>> http://cr.opensolaris.org/~jmellors/core-ips-extpkgs/ >>> >>> Agents: >>> http://cr.opensolaris.org/~jmellors/ds-ips-extpkgs/ >>> >>> Thanks >>> Jonathan >>> _______________________________________________ >>> ha-clusters-discuss mailing list >>> ha-clusters-discuss at opensolaris.org >>> http://mail.opensolaris.org/mailman/listinfo/ha-clusters-discuss
