Hi Dave,

Thank you for the code review.  Please see my response to your comments 
below.

On 01/15/10 07:44 AM, Dave Miner wrote:
>
> ai_sparc_image.xml (and the other manifests): I'd suggest the new 
> parameter closer to the top, before the "rarely configured" signpost 
> (and probably before the package list).
I have moved the root password parameter to before the package list for 
all the manifests.
>
> pre_boot_archive_pkg_image_mod.py:
>
> I'm bothered that we're proliferating mechanisms for setting 
> passwords.  ICT presently uses the PasswordFile class from IPS to do 
> essentially the same thing (minus the encryption).  I'll leave it to 
> you as to whether you want to reconsider this now or file a bug to 
> deal with it later; comments below in case you stick with this code:
>
Using the PasswordFile class from IPS is a great idea.  I updated the 
code to use that instead of
writing my own.

> 105: Can break from loop at this point
>
> 120: we're not saving timezone here; seems this line should be deleted
>
> install_utils.py
>
> 770: any reason not to just use string.printable?
The reason I can not use string.printable because characters from the salt
should only contain characters from the set [a-zA-Z0-9./] according to 
the crypt_unix(5) man page.

An updated webrev with the changes I made addressing your comments and 
Keith's comments
is here:

http://cr.opensolaris.org/~ktung/12726-updated/

Let me know if you have further comments.

Thank you again for your review.

--Karen


Reply via email to