approved in 7/23 meeting [Re: RaidCfg GUID and volume activation support [PSARC/2008/458 FastTrack timeout 07/24/2008]]

[Tzongyu Paul Lee]

Yu-Bo Ryan Wang 写道:
> Hi James,
>
> Please see my in-line comments.
>
> Thanks,
> Ryan
>
>
> James Carlson wrote:
>> Yu-Bo Ryan Wang writes:
>>   
>>> For the output change, as the section 4.3  mentioned,
>>>
>>>         GUID information output had been added to disk information display
>>> 	in a compatible format. And for a given volume, if it's in inactive
>>>         state, array information display output will show this status. See
>>> 	accompanied diff file for the raidctl manpage.
>>>
>>> Currently, there is nothing dependent on that output.
>>>     
>>
>> I'm a little surprised that adding a field in the middle of a table is
>> considered a "compatible format," particularly so in a patch, but if
>> we're absolutely sure that there's nothing dependent on this but human
>> eyeballs, then it seems ok to me.
>>   
> We are sure that there's no dependent on this because raidctl is a 
> config tool
> for RAID and it's not a frequently used software; and according to our 
> survey
> during the research period, the only possible exception (besides human 
> eyeballs)
> is the exit status checking, and this is un-changed.
>> Please do make sure that when the man page is updated, the stability
>> is updated to the current terminology (what was once "Evolving" is now
>> "Committed"), and that the man page makes clear that the stability
>> applies to the command line arguments only, as the output from this
>> command (as this project makes plain) is substantially less than
>> "Committed."  It's probably "Volatile" at best.
>>
>>   
> Agree. We will do this when updating the man page.
I suggest to file a CR/RFE to capture this so that this will be 
reflected when you next do any
putback related to this.


-- 
Tzongyu Paul Lee, Tzongyu.Lee at Sun.Com or Paul.Lee at Sun.COM
BJS05 7225, x84343