On Wed, 2008-11-19 at 12:22 -0500, Peter Memishian wrote:
> > 
>  > zhadum(/export/ws/clearview/clearview-ipmp)$ hg heads
>  > changeset:   8081:fd7ae55e589c
>  > tag:         tip
>  > parent:      8079:4a841b81edb3
>  > user:        meem <Peter.Memishian at Sun.COM>
>  > date:        Wed Nov 19 01:18:23 2008 -0500
>  > description:
>  >         Clearview IPMP implementation: renames and additions
>  > 
>  > changeset:   8080:4c3a4bf090cc
>  > user:        meem <Peter.Memishian at Sun.COM>
>  > date:        Tue Nov 18 18:06:03 2008 -0500
>  > description:
>  >         makefile formatting
>  > 
>  > On a related note, in order for the children of clearvew-ipmp to go
>  > through this exercise successfully, the new clearview-ipmp and the old
>  > clearview-ipmp(-ipobs) must have identical file contents, otherwise the
>  > children risk undoing changes that have been made to the new
>  > clearview-ipmp (was the "makefile formatting" chageset in the old repo
>  > as well?).
> 
> No, the Makefile formatting thing is the issue you asked me to resolve
> in Makefile.rules.
> 
> I see what you're saying above, but that seems to throw a wrench in this
> whole thing, as it means I cannot integrate any progress into the IPMP
> gate until Erik and anyone else with children complete their reparent.
> This is quickly turning into a mess, and I'm still not sure what problem
> we were trying to solve by compressing these changesets out :-(

I'm starting to feel the same way, but we're already rather far down
this road. :-(  Perhaps we can simply take a clone of clearview-ipmp
that is based on the changeset that is equivalent to the tip of
clearview-ipmp-ipobs, and use that for the transition.  That way, you
can keep working off of clearview-ipmp, and we can reparent to that once
the ipobs changesets have been blown away in the children.

-Seb



Reply via email to