On 7/28/12 5:22 AM, Hanspeter Niederstrasser wrote: > On 7/27/2012 6:57 PM, Alexander Hansen wrote: >> + echo "I am not overwriting this." >> + echo "You will need to edit $PREFIX/etc/passwd-fink manually" >> + echo "and set the UID entry for $FINK_PASSWD_USER to an unused" >> + echo "value." > > Since apparently numbers below 500 are now taken to be for Apple's use > only, should the instructions say to pick a number above 600? And maybe > instructions on how to make sure that hand-picked number is not being > used? The command "dscl . -search /Users uid <ID_TO_USE>" outputs an > entry if it exists (blank otherwise). > > Suggested rewording: > > ... and set the UID entry for $FINK_PASSWD_USER to an unused value above > 600. Run the command > > dscl . -search /Users uid <VALUE_TO_TEST> > > to verify the value you chose is currently not in use. > > ****** > > Since this command returns an empty value when unset, maybe we can write > a perl script that starts searching at 600 until it finds an unset value > and then output that as a possibility. This has the benefit of thinking > ahead if we want to do automatic UID/GID assignment w/out manual file > editing. >
I'll see if I can do it in the existing shell script. > Finally, may I also suggest: > > 1) modify passwd-core InstallScript as below. This adds README to > passwd-core so that split out passwd-* packages can have a file for > their deb. Needed if #2 below is done, otherwise it is just a nicety. > > InstallScript: << > mkdir -p %i/etc %i/sbin %i/share/doc/%n > install -c -m 644 passwd-fink %i/etc/ > install -c -m 644 group-fink %i/etc/ > install -c -m 755 update-passwd %i/sbin/ > install -c -m 644 README %i/share/doc/%n > << > Sure. > 2) move the passwd-* packages into individual files. This way they can > be updated individually when needed w/out affecting other passwd-users. > Once the backend is updated w/ whatever changes are deemed necessary, > I'm happy to do the manual labor of doing the actual new .info file > creation. > That sounds good. I'll leave the prior version around. > 3) Should passwd-dovecot be erased, or made obsolete like fink-bld was? > It should probably be made obsolete, yeah. I just erased it because it wasn't likely to be installed on 10.7 due to the lack of dovecot. :-) On 10.5/10.6 we'll actually need to do some testing, since 10.5 lacks a "_dovecot" user, and even on 10.6/10.7 there's not a "dovecot" alias that points to "_dovecot" (however DirectoryServices handles that). > Last note: the attached passwd-core.patch file includes an already > parsed passwd-fink, as well as passwd-fink.in that is needed (where > dovecot is not commented out). > > Hanspeter > Oops. -- Alexander Hansen, Ph.D. Fink User Liaison My package updates: http://finkakh.wordpress.com/ ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
