Danek,
Thanks for looking at this.
Danek Duvall wrote:
On Thu, Mar 12, 2009 at 04:02:13PM -0600, Jerry Jelinek wrote:
http://cr.opensolaris.org/~gjelinek/webrev.7304/
In the second hunk, why bother testing at all? You're discarding all the
output of the zfs create, so why not rely on that when the create fails
because the dataset already exists? All you'd need to change is to add the
redirection to /dev/null on the create command that exists.
Well, thats basically what we were doing minus the redirection.
I thought maybe it would be a little more efficient to check first,
but maybe that was a bad idea and I don't have any data on that one
way or the other. I can go back to what we had plus the redirection.
I'll go ahead and ask why you don't do that for the first hunk as well,
though the answer I expect is that because you actually want to see the
message when it fails. Except that you also print out $f_zfs_create, so
I'm not sure.
Yes, I think we should fail if we really can't create the dataset and
seeing the actual zfs error msg might help with troubleshooting in this
case.
Thanks again for reviewing this, I'll make the change from above,
Jerry
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
