But it's the pkg install --reject option that it most mimics. On it's own the reject action will do nothing useful, it has to be used in conjunction with an install action.
See you suggesting it should do something else? Thanks, Darren. Ethan Quach <[email protected]> wrote: > > >On 03/20/12 08:14, Darren Kenny wrote: >> >> On 20/03/2012 14:50, Dave Miner wrote: >>> On 03/20/12 10:42, Darren Kenny wrote: >>>> On Tue Mar 20 14:14:33 2012, Dave Miner wrote: >>>>> On 03/20/12 04:10, Darren Kenny wrote: >>>>>> Hi Shawn, >>>>>> >>>>>> On 16/03/2012 20:12, Shawn Walker wrote: >>>>>>> On 03/16/12 05:38, Darren Kenny wrote: >>>>>>>> Hi, >>>>>>>> >>>>>>>> Could I please get a code review for the following bug and RFE: >>>>>>>> >>>>>>>> 7152537 AI fails to install packages with licenses using >must-display=true >>>>>>>> and not must-accept=true >>>>>>>> >>>>>>>> 7145997 noinstall element should be implemented as reject for >>>>>>>> pkg transfer >>>>>>> It may be too late to change this, but 'noinstall' doesn't >exactly roll >>>>>>> off the tongue. Why was this term used instead of 'reject'? >>>>>> At the time that the DTD was first envisaged, the term 'reject' >wasn't in >>>>>> use by pkg. >>>>>> >>>>>> Other than that, I've no idea why noinstall was used originally. >>>>>> >>>>>> We could look at a bug to change it at a later date if the DTD is >being >>>>>> reved anyway - but otherwise we are pretty much stuck with it, >even revving >>>>>> it could actually present reason for maintaining support for both >names... >>>>>> >>>>> Since we've not previously implemented this for any of the >transfer >>>>> modes, I think it could be changed without concern for >compatibility. >>>>> "Reject" isn't a term that is commonly used with other modes such >as >>>>> cpio, but then neither is "noinstall". I'd probably opt to change >it >>>>> based on the assumption that most of the uses that end-users will >have >>>>> would be pkg-style transfers, though I'm planning to implement it >for >>>>> cpio internally at least (7123561). >>>> OK - so you're saying it would be OK to change this without revving >the >>>> DTD? Based on no existing users that makes sense. >>>> >>> Right. >>> >>>> So could I propose then that we use the term "exclude" - this would >>>> seem to >>>> fit the pkg and CPIO usages better? >>>> >>> I'd be more inclined to ensure we had terminology that aligned with >pkg >>> in this case. I could see having both as synonyms, I guess, but I'm >not >>> sure it's worth the trouble. >>> >> Ok, so "reject" it is then... > >Should it be reject or avoid? The latter has its own pkg subcommand >which would seem to make it more aligned with our install and uninstall > >action types for processing ips packages. > > >-ethan > >> >>> We do need to file a small ARC case to record whatever change. >>> >>> Dave >> _______________________________________________ >> caiman-discuss mailing list >> [email protected] >> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >_______________________________________________ >caiman-discuss mailing list >[email protected] >http://mail.opensolaris.org/mailman/listinfo/caiman-discuss _______________________________________________ caiman-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
