Danek Duvall wrote:
On Thu, Aug 13, 2009 at 10:25:02AM -0700, Danek Duvall wrote:
On Wed, Aug 12, 2009 at 09:34:22PM -0700, Danek Duvall wrote:
http://cr.opensolaris.org/~dduvall/pkg-bug10630/
I'll probably update this webrev tomorrow with a test case, too.
So updated.
And updated once more with a new message that illuminates what BE the
rem_drv should be performed on.
Jan had the suggestion that we could look at the original driver_aliases
file we ship (i.e., pull it from the download cache), and assume that any
conflicts we have with aliases registered there can be overridden silently,
while other conflicts can be assumed to be there because of an explicit
user action, or from an third-party SVr4 package.
I think that's a good idea, despite the almost certain grottiness of trying
to extract the original file,
The trade-off is probably what level of gross code we're willing to
tolerate in order to support transition from the earlier opensolaris bits.
but too much work to make it in for this
particular bugfix, and now that we're starting work on getting all drivers
represented by driver actions, much lower priority.
I'd consider just pulling in a hard-coded list when we turn everything
into actions. That way we know it's the final set, and it simplifies the
special case code somewhat.
-jan
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
