Hi Sundar.
Thanks for the clarification.
Sundar Yamunachari wrote:
> Ethan Quach wrote:
>>
>>
>> Sundar Yamunachari wrote:
>>> Jack Schwartz wrote:
>>>> Hi Ethan.
>>>>
>>>> Ethan Quach wrote:
>>>>> Jack,
>>>>>
>>>>> 2.2
>>>>>
>>>>> If the SC manifest is an enhanced smf profile then I don't see a
>>>>> need for the install tools (our parser) to understand how to
>>>>> parse and validate them. We've requested from the smf team
>>>>> the ability to use svccfg(1M) to validate a profile, which will
>>>>> do basic syntactical validation. We don't own the form of this
>>>>> file, so beyond that I don't see much more we could or should do.
>>>> I also asked about this when given the request, but our SMF enhance
>>>> profiles team said they wanted it. I didn't know about the svccfg
>>>> validation request from us. CC'ing Sundar, Evan and Joe to get
>>>> their response on this.
>>> We had a similar discussion when Jack sent the strawman proposal. I
>>> have attached the message. I have send this requirement just to
>>> capture that we need a parser to validate the enhanced SMF profile.
>>
>> Let me try to clarify what you're saying here: you're saying you sent
>> out the requirement just to note it, and if what's doing the validation
>> comes from something else, like svccfg, then that fulfills your
>> requirement.
>>
>> Is that correct?
> Yes. I wanted to make sure that there is a way to validate SC
> manifest. If it comes from svccfg, that is fine with me.
OK. I am changing the assumptions in the document, removing the
requirement that the XML parser can handle DTDs and validation of
enhanced SMF profiles.
The doc will be changed as follows:
- - -
2.2 Assumptions
SMF enhanced profiles will be used to do system configuration. These
files will be validated outside a parser (likely via svccfg).
A parser will be picked which can understand and validate against
RelaxNG schemas.
- - -
Thanks,
Jack
