Darren, I'll leave that to you to prove it's okay within Sun - I can only state what I have seen on previous reviews/RTIs. But unless you can, I can't okay your review - sorry :-( I would have given you a name to ask, but the axeman came and chopped.
paul Darren J Moffat wrote: > Paul Cunningham wrote: >> >> >> Amanda Waite wrote: >>> >>> As for the 755 perms, you're right in that if makes sense in this >>> instance to use the permissions of the files generated by the build. >>> I'll add that to my list of things to consider when reviewing webrevs. >> >> I think the rule (now) might be that all new files in /usr should >> *not* have any write permission bits set (unless there is a real >> reason to). > > Which ARC case has this rule written down ? > > I'm looking and asking around to see if there was ever an ARC case that > set precedent in this area. > > The closest I've found so far is a half-pager from Casper Dik > (http://sac.eng/Archives/Projects/1998/19980826_casper.dik sorry there > isn't an opensolaris.org version of this old case). > >> I've seen recent RTIs be rejected because they are set. > >> But maybe someone in the SFWNV C-team needs to confirm that (if there >> are any members of that team left that is). > > I believe ON has exactly the opposite rule, ie all our binaries and libs > that are root owned are owner writeable. > > I'm trying to confirm this just now. > >> I personal never understood why anything in /usr needed any write >> permission bits set. > > So installation can work properly. > -- ---------------------------------------------------------------------- Paul Cunningham Software Engineer Tadpole Business Unit
