Nut upsdev - Jul 2005 - newhidups, who clears ups_status?

Gentlemee,

I have just encountered a problem with low battery status in newhidups. 
Here is the situation:
1) The system goes through the shutdown sequence due to low battery,
   the UPS receives the shutdown.return and shuts off power,
2) After power returns, the UPS turns on power before its battery
   reserves have reached low battery level (a potentially risky 
   situation if we get another power failure immediately)
3) The UPS therefore reports low battery status to newhidups for the
   first few cycles (actually maybe just once!), 
4) The UPS reaches enough reserves and no longer reports low battery

Question: Looking at process_status_info(), ups_status seems to be used 
cumulatively, and alternative states clear each other, but who clears 
low battery from ups_status?

I have observed upsc report a status of OL+LB+CHRG on a well charged 
battery several times, and I have just accidentally created a short power 
interruption for a fully loaded UPS and upsmon went immediately into a 
shutdown sequence despite the 100% charge! 

Sorry, but I think something is broken, AG

-- 
 ----------------------------------------------------------------------
   Alfred Ganz					alfred-ganz@agci.com
   AG Consulting, Inc.				(203) 624-9667
   440 Prospect Street # 11
   New Haven, CT 06511
 ----------------------------------------------------------------------

[@Alfred: I keep cc'ing you as you leave your @ before...]
> I have just encountered a problem with low battery status in newhidups.
> Here is the situation:
> ...
uh, good catch.
I've forgotten to complete the new behavior with
the new process_status_info(), clearing LB when either
UPS.PowerSummary.PresentStatus.ShutdownImminent
or
UPS.PowerSummary.PresentStatus.BelowRemainingCapacityLimit
is set with value "0" to reset

Corrected in the CVS Development branch.
Thanks to test and ack back.

Note that I've put ShutdownImm before BelowRCL in
*-hid.h files as (on MGE at least) the LB flag would
be cleared where it should not...

Arnaud