Hi Jack,

Many thanks for your feedback :-) I follow most of what you are saying and it makes sense to me. One thing I would like to clarify with you is that an older AI client in particular will trip up if there are gpt_partition elements in the manifest it tries to process. If not first caught by the manifest parser, then the AI client itself will surely fail as soon as it tries to parse the manifest data if gpt_partition elements are defined. The items are not required however as you note, but S11 FCS AI will not be able to handle them if present - the newer, GPT aware, AI client would be required.

Thanks,
Niall

On 12/31/11 06:32 AM, Jack Schwartz wrote:
Hi Niall.

I can see the need to rev the DTD version if it specifies new *required* items for a manifest. But as long as the new items here are not required in a manifest, old manifests will still validate with the new DTDs. Given this, then as long as AI and DC can differentiate and handle both GPT and non-GPT cases I think the version can stay the same.

    My $.02,
    Jack

On 12/30/11 05:59 AM, Niall Power wrote:
Hi,

During slim_uefi code review Darren suggested that we might need to bump the AI and possibly the DC manifest DTD versions because we have bumped the target DTD version which both AI and DC include.

For reference, the webrev including the target.dtd changes is here:
https://cr.opensolaris.org/action/browse/caiman/niall/slim_uefi_version_01/webrev-uefi-ver-1/

Here are a few factoids:
- the target.dtd change is restricted to the physical section of target so there would be no practical or visible impact to DC users - the new version of target.dtd is backwards compatible with the previous version of target.dtd. We only add in definitions for GPT elements. The existing elements are unchanged (minus a bug fix that makes partitions. slices and gpt partitions mutually exclusive
  direct children of a disk)
- An AI manifest that defines GPT partitions would fail to validate against an AI client that only expects non-GPT target elements ie. slices and MBR partitions.

Given the above, and other considerations that I am sure I have overlooked, I'd appreciate some feedback or guidance on what the appropriate course of action might be.

Thanks!
Niall
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss


_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss

Reply via email to