The current behaviour of pkg(1) when attempting a 'pkg install' of a
package with "reboot needed" on the live image is nice and safe and
exactly what I would want to happen on a production system (and even on
my desktop & laptop).
However in a development environment where I'm constantly updating the
same package because I'm testing fixes (or worse resorting to printf
debugging in some cases) I don't really need an additional 10+ entries
in my grub menu and beadm output each day (some days many more than that).
I believe I know how to implement a "--force" for 'pkg install' that
would allow the update on the live image, but before I code it up I want
to be sure that it would be accepted into the gate. I know that in
production use this could be dangerous and is providing a means to by
pass the safety net provided by using boot environments. So maybe it
could be undocumented or print a big warning ?
I could just avoid using pkg(1) and cp the binaries into place (like bfu
used to do) but I want to use the tools and help test them.
--
Darren J Moffat
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
