Hi Keith,

please see my response in line.

Thank you !
Jan


Keith Mitchell wrote:
> Hi Jan,
>
> Responses below. I removed anything I have no further comments on. 
> Thanks!
>
> - Keith
>
> Jan Damborsky wrote:
>> Hi Keith,
>>
>> thank you for your comments - I have accepted almost
>> all of them - please see below.
>>
>> I have updated webrev accordingly - it now contains only
>> changes related to your comments:
>>
>> http://cr.opensolaris.org/~dambi/python-24-26/
>>
>> Thank you,
>> Jan
>>
>>
>>>
>>>
>>> 401-402: This line would be more readable if split up.
>>
>> I am not sure how you would like to have it split up. Could you
>> please elaborate more on this ?
>>
>>
>
> Something along the lines of the following would work:
> svc_address = service_list[svc_found_index].get_txt_rec()
> svc_address = svc_address.strip().split('aiwebserver=', 1)[1]
> svc_address = svc_address.split(',')[0]
> svc_address, svc_port = svc_address.split(':')
>
> It still looks somewhat messy, but it breaks up the chain of method 
> calls so they're not wrapping multiple lines.

I have changed this.

>
>>
>>> 216-224: The 'platform' python module provides this functionality 
>>> internally. Alternatively, os.uname() also exposes this data.
>>> Same comments for code in AICriteriaPlatform and AICriteriaCPU
>>
>> Well, I would prefer not to change this part at this point - also
>> I haven't found equivalent for 'uname -i' - do you happen to know
>> if there is one in 'platform' or 'os' module ?
>
> Ah, you're correct here, there doesn't appear to be a '-i' equivalent. 
> The module non-public function "platform._syscmd_uname('-i')" appears 
> to be a shortcut for running uname, but as it's a non-public 
> interface, we can't rely on its stability.

Agreed. Based on this, I didn't modify that part.
I have pushed all the changes into 2.6. gate.


Reply via email to