Sanjay, Thank you for the help and feedback!
My comments regarding the "C" code are in line below. William will address your Python feedback. Updated webrev will be sent shortly. Joe Sanjay Nadkarni wrote: > Here are my comments. > > > b/libict/Makefile: > Are ROOTADMINLIB still required ? ROOTADMINLIB is pervasive in the makefiles. I have discussed this with Moriah and I think it is beyond the scope of this project to determine it's need. I have filed bug Bug 3599 Evaluation and cleanup should be done on the /lib Makefiles. http://defect.opensolaris.org/bz/show_bug.cgi?id=3599 > > do we need to build static libs ? no. removed > > #define ICT_STR_ERROR > I am frankly not in favor of big macros. They are hard to maintain over > time. Since nothing here is performance critical a big switch statement > might be nicer. As we discussed this is a style issue. I will leave it as it is. > > ict_create_user_directory: > It should be zfs create .... i.e. a dataset > That is what is being recommended. fixed > > ict_snapshot - nvlist is not freed after successful snapshot. fixed - good catch! > ict.py > 103 #import dumper ? > > 841: zvol can be a dump device > http://docs.sun.com/app/docs/doc/817-2271/6mhupg6nk?l=ru&a=view > Note that the block size should be set to 128k > (this impacts ti too..) > > Should add_sysidtool_sys_unconfig_entries > be more flexible. i.e. I could forsee a need that other installers may > need to either or change the applications. > > should remove_install_specific_packages have the ability to accept a > list of pkgs. > This list may be different for different installers. For example, AI > may not have SUNWslim-utils. > > 1334 cleanup_unneeded_files_and_directories. > I think it should be able to accept additional files. > > failsafe related code can be removed. > >
