On  6/05/10 05:56 AM, Mark J. Nelson wrote:
Please review

        http://cr.opensolaris.org/~mjnelson/webrev.moredeps/

This patch, though nonintersecting, is applied on top of and tested
along with this one, which was previously reviewed:

        http://cr.opensolaris.org/~mjnelson/webrev.deps/

Most of the changes are package dependencies, but I'm hoping somebody
with decent pkg build fu will review the usr/src/pkg/Makefile changes.

I don't claim to have such fu, but what I see there (as well as the
other changes) does look reasonable.


Specific questions on those bits:

1. I made the desktop-mime-cache restart_fmri a universal transform.  I
believe that's appropriate, but would like confirmation.

I concur.

2. I temporarily enabled dependency generation and resolution on the
onbld tools package.  This pulled in the following dependencies:

depend fmri=pkg:/[email protected],5.11-0.140 type=require
depend 
fmri=pkg:/developer/versioning/[email protected],5.11-0.137:20100409T144402Z 
type=require
depend fmri=pkg:/library/[email protected],5.11-0.137:20100409T151317Z type=require
depend fmri=pkg:/package/[email protected],5.11-0.137:20100409T151531Z type=require
depend fmri=pkg:/runtime/[email protected],5.11-0.140 type=require
depend fmri=pkg:/runtime/[email protected],5.11-0.137:20100409T152125Z 
type=require
depend fmri=pkg:/runtime/[email protected],5.11-0.137:20100409T152239Z 
type=require
depend fmri=pkg:/system/[email protected],5.11-0.140 type=require

I then added these dependencies, minus the version information, to the
onbld manifest.  If anybody can think of a better way, I'm happy to hear
it.  But using pkgdepend gets package versions, which causes the
incorporation to be updated.

I'm fine with what you've done

3. The parallel publication changes mean a build machine flag day to
require build 137.  Is it reasonable, therefore, to rip out the
conditional setting for SUPPRESSDEPS, as I did?  Or are we still
expecting people to be bfu'ing merrily along?  I guess the difference
would be "we now require build 137" vs "we now require OpenSolaris build
137."


The only issue I've got with requiring snv_137 for build machines
is that people outside the firewall are going to be in a bit of a
hole until 137 or later is available on pkg.opensolaris.org/dev.

I do, however, like the idea of bringing build machines up to a
newer build in general.


overall +1 from me.


James C. McPherson
--
Senior Software Engineer, Solaris
Oracle
http://www.jmcp.homeunix.com/blog
_______________________________________________
on-ips-dev mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/on-ips-dev

Reply via email to