Hi Jack,
On 11/ 1/10 06:45 PM, Jack Schwartz wrote:
> Hi Darren.
>
> I like how this doc keeps track of the many details that AI needs to manage.
>
> Here are my comments.
Thanks for your comments.
>
> General comment: Derived Manifests is not mentioned anywhere. While
> not part of the AI->CUD project, it is part of AI. It will be a
> checkpoint, and other checkpoints not part of AI->CUD are mentioned in
> the document.
I added a reference to it as an imported interface, dependency, and in section
3.3.1.1 to describe what we will need to do with it.
>
> Page 7, top bullet, last sentence: can be clarified as "This project
> will fully implement the new manifest design."
Done.
>
> Nit: Section 3.3, 1st bullet, second sub-bullet: creating -> create
Done
>
> Nit: section 3.3, second bullet: starts -> start
Done.
>
> Nit: Page 8, top: Isn't "Run the Install" the same as "STart the engint
> to perform the installation" of the previous bullet?
Yes, removed that piece of text.
>
> Page 9: bottom: <add_drivers> has a <software> subtag as well as a
> <search_all> subtag. <software> elements are used to explicitly specify
> (driver) packages to install.
Added that.
>
> Page 10, end of section 3.3.1.3: Is another bullet for "other tags
> handled by checkpoints" appropriate here?
I'm not sure I want to list all the tags that other checkpoints handle - this
should really be part of their design. I only included the DDU tags because they
are unique to AI, and we will be creating the new checkpoint as part of this
project.
>
> Nit: 3.4.2: middle bullet: There are 2 periods at the end of the sentence.
Fixed.
>
> Page 11, bullet near top: Same comment about <add_drivers> having a
> <software> subtag as well as <search_all>.
Done.
>
> Nit: 3.4.3. last sentence: "can them be" -> "can then be"
Fixed.
>
> 3.4.5 Is zpool is more appropriate than pool here?
Yes. fixed.
>
> Nit: 3.4.6, sentence below the "/persistent/desired" box:
> "apply it to the system for real" -> "modify the system"
> then start the next sentence with "Note: this is the first point..."
Done.
>
> Nit: get rid of the "etc" bullet atop page 13.
Done.
>
> Page 14, middle: Under the "Expected, but not yet planned" section: I
> don't understand what "Adding of entries for sysidtool and sys-unconfig
> to run all known external applications" means and pertains to. Can you
> please explain?
Honestly, I don't know - this is just a reference to the existing ICT in
usr/src/lib/libict_pymod/ict.py, which :
creates /etc/.sysidconfig.apps
touches /etc/.UNCONFIGURED
copy .sysIDtool.state to the target
I don't know why this is necessary.
>
> Nit: page 15, grammatical changes to sentence above the bullets near the
> bottom:
> "For boot configuration, the ... required. Where something ...
> platform, its (not it's) implementation... just be empty and do nothing."
Switched things around.
>
> 3.5.2: AI will be using the Device Driver Utility Common Library
> interfaces, PSARC/2009/602
Added this.
Thanks,
Darren.
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
