Jonathan Mellors wrote: > Hi Nick, > > I think it would be better if we moved the EXTPKG variant entries to the > top of the rule after $(RM), like for the normal IPSMANIFEST rule, so if > anyone looks directly at the repository manifest, it's easy to see which > platform it installs on. And we'd also have to do the import after the > include command (Core 95,96; Agents 93,94) to achieve that. >
Thanks, Jonathan. I've made your suggested change and removed the zone variant line because it turns out it wasn't actually doing anything. See http://www.mail-archive.com/pkg-discuss at opensolaris.org/msg11514.html and the rest of the messages in that thread for details. Revised review: http://cr.opensolaris.org/~nsolter/variants-core-v2/ http://cr.opensolaris.org/~nsolter/variants-ds-v2/ Thanks, Nick > Thanks > Jonathan > > Nicholas Solter wrote: >> Jonathan, et. al., >> >> I've updated the build to add two variant lines to each IPS package: >> >> set name=variant.arch value=$MACH >> set name=variant.zone value=global >> >> These prevent the packages from being installed on the wrong platform, >> or in non-global zones. >> >> I've also added this line, though I think it's ignored at the moment: >> >> set name=info.classification >> value="org.opensolaris.category.2008:System/HA Cluster" >> >> The code changes are minimal for now. When we support non-global >> zones, we'll have to customize the zone variant line for each packge. >> >> Core: >> >> http://cr.opensolaris.org/~nsolter/variants-core/ >> >> Agents: >> >> http://cr.opensolaris.org/~nsolter/variants-ds/ >> >> Thanks, >> Nick >> _______________________________________________ >> ha-clusters-discuss mailing list >> ha-clusters-discuss at opensolaris.org >> http://mail.opensolaris.org/mailman/listinfo/ha-clusters-discuss
