On Thu, Jul 23, 2009 at 06:27:05PM +0100, Alan Maguire wrote:
> thanks for the comments! responses inline...

Sounds good, thanks.

-renee

> Renee Danson Sommerfeld wrote:
>> Hi Alan,
>>
>> Just a few comments.  The first two are really code nits; the third
>> is a point I'm not clear on.
>>
>> Thanks!
>> renee
>>
>> nwamadm.c: rather than moving the check of the nwam service state out
>>     to the individual functions where it matters, could you make the
>>     check in main() conditional?  That is, have a set of subcmds that
>>     require nwam to be online, and continue to do the state check in
>>     main() if the requested subcmd is in that set.
>>
>>   
> sure. I could add a boolean to the cmd structure
> that specifies if the command requires nwamd
> to be running.
>> libnwam.h'57: not sure why this change was needed?
>>
>>   
> oops. I had thought I'd need another flag,
> turned out I didn't.
>> libnwam_loc.c'loc_set_enabled(): It looks like we just enable/disable
>>     as needed while walking the list of locations.
> Yep.
>>  What happens if
>>     we hit the to-be-enabled location before other, currently enabled
>>     locations?  Does this mean there will be a window where multiple
>>     locations are enabled?  What's the consequence of that?
>>
>>   
> The first-encountered location is enabled, it won't
> be an issue for the other location since as
> part of the enable we do a request to nwamd to
> enable the to-be-enabled location. Then the
> to-be-disabled location will be marked as enabled=false.
> So in other words, no negative consequences
> as far as I can see.
>
> Thanks!
>
> Alan

Reply via email to