apcupsd- 3.10.18-3

[Orion Poplawski]

CCing extras-list to make public.  Perhaps bugzilla would have been the
appropriate place.

Luboš Staněk wrote:
> Hi Orion,
> I am using apcupsd on my machines due to poor USB handling by nut.
> 
> I have reviewed your changes to the extras package and I consider them
> insufficient.
> 
> Problems:
> 
> 1) init script:
> - Reload action: do you think that reload and restart are the same? I do
> not! I think that the reload means that the application can handle
> config reloading after a signal (frequently SIGHUP) without breaking
> current status. Consider that master reloads itself your way but all
> slaves may loose connections.
> - Autostarting the daemon without proper configuration is meaningless if
> you do not provide autoconfiguration mechanism.

Other services that don't handle SIGHUP like methods have reload mapped
to resart (e.g. dhcp).  I guess the question is what is worse - having
"service apcupsd reload" not do *anything*, or have it do a restart.

Service is not started automatically.

> 2) SMP: I am working on AMD Athlon 64 X2 machine and the -j2 build works
> fine.
> 

-j 3 broke on at least one machine so it's gone.  Besides, that's just a
small build time issue.

> 3) Your package does not build and distribute the USB driver.
> 

Not sure I follow.  I have no trouble talking with USB UPSes.  hid-ups
is just a test program, not a driver.

> 4) Patching and modifying runtime initscripts without reapplying changes
> after their update is not good. Either ask initscripts maintainers to
> add apcupsd handling in the halt script or use other solution.
> 

Where is this being done?


-- 
Orion Poplawski
System Administrator                   303-415-9701 x222
Colorado Research Associates/NWRA      FAX: 303-415-9702
3380 Mitchell Lane, Boulder CO 80301   http://www.co-ra.com