[pkg-discuss] Code review request for build 175a resync

[david...@oracle.com]

I'd appreciate a review of the following webrev

https://cr.opensolaris.org/action/browse/pkg/comay/Build_175a_Resync/webrev/

which covers the changes for the following

18986 crypto/gnupg should not be in solaris-large-server
19002 resync repository to snv_175a
_______________________________________________
pkg-discuss mailing list
pkg-d...@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

[Danek Duvall]

david...@oracle.com wrote:

> https://cr.opensolaris.org/action/browse/pkg/comay/Build_175a_Resync/webrev/

set-publisher.sh:

- How invasive would the changes to other things be to make -b take the
entire branch, and not something that will have "0." prepended?

- line 116: I'm not sure I understand why you changed the way this
worked. Why not simply use a colon as the terminator? Are you trying
to allow for any subsequence of the given branch?

- line 129, 141: what are these echos for?

Danek

[david...@oracle.com]

> - How invasive would the changes to other things be to make -b take the
> entire branch, and not something that will have "0." prepended?

It's certainly doable but I'd prefer at this point to leave this as-is
since there are quite a few *_BRANCHID environment variables and I'm
concerned about the impact on RE. Also, I expect (hope?) all of this
will be retired sooner rather than later.

> - line 116: I'm not sure I understand why you changed the way this
> worked. Why not simply use a colon as the terminator? Are you trying
> to allow for any subsequence of the given branch?

I made this change since we want to match the precise branchid modulo
the nightlyid. So rather than having to quote the periods with
egrep(1), I switched to fgrep(1) and using ksh's {var%.*} to drop the
nightlyid.

> - line 129, 141: what are these echos for?

Mostly to make the set-publisher.sh output be a bit clearer since each
consolidation's output runs together.

Does that answer your concerns? And thanks for the review.

[Danek Duvall]

david...@oracle.com wrote:

> Does that answer your concerns? And thanks for the review.

Yup, thanks.

Danek