On 10/06/11 14:01, Danek Duvall wrote:
Shawn Walker wrote:
On 10/06/11 13:43, Danek Duvall wrote:
Shawn Walker wrote:
https://cr.opensolaris.org/action/browse/pkg/swalker/pkg-bttf/webrev/
elf.py:
- Why /usr/ccs/lib and /usr/sfw/lib? These aren't in the runtime
linker's default runpath.
Because unit tests fail otherwise on 151.
Could you go into a bit more detail on why? As far as I can tell, those
paths were never in that file, so how would the unit tests have passed
previously?
The api.t_dependencies tests were failing. But I see now that was
because I was using gcc to build the gate originally when I made that
change, and gcc definitely brings those paths into play.
So I'll undo this change.
ips-incorporation.p5m:
- Why is cherrypy different than the others?
Just missed it when I moved the rest down a few lines.
And it gets the pkg.debug.keepme=true attribute?
Yes, it will. Thanks for spotting that.
The brand scripts are from changeset 2219 of the gate just before the
dataset changes Mike made.
Be careful -- not everyone's 2219 is the same, and mine very definitely
isn't what yours is. Thankfully, your diff included the changeset id,
which let me figure out what was going on.
I'm confused as to how it couldn't be the same 2219 unless you had a
workspace that was really old and had changes merged around from that
point. But yes, 2219:60ad60f7592c specifically.
Hm. Okay, that's the same 2219 I have, but it doesn't precede Mike's
changes. The diff header you had showed a diff against 33dd51b5ca54, which
*is* the one right before Mike's changes. So now I'm confused about what
you were going for, unless "just before" can be stretched to "16 changeset
before".
Yes, sorry, when I did the 'hg revert' or 'hg up' originally I used
changeset 2234 for the brand scripts. But the brand scripts actually
last changed in 2219. That's what I get for mixing log output up.
-Shawn
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
