[Nut-upsuser] Problem with APC and Fedora 8 I86_64

[Geraint Evans]

Thank you and Arjen for the prompt reply.  I have tried to work back through this to give you some 
further information but I am storm dodging, over here, and the Fedora repository has been going off 
line so it may take a while :( . Sorry, I'm not a developer and I thought that the redirect on the 
commands would redirect the warnings as well, I will retry so I capture the stderr stream as well. 
Unfortunately this will be a production machine so I cannot volunteer any 64 bit testing time on 
it. It looks like I am going to have to shut down for the rest of the day due to more storms (a 
side effect of living in the tropics) but I will try the repositories again tomorrow so that I can 
get back to you with more details.

G Evans


On Fri 15 Aug 2008 18:56:54 Charles Lepple wrote:
> On Fri, Aug 15, 2008 at 4:48 PM, Geraint Evans <pvcymraig at yahoo.co.uk> wrote:
> > I have been trying to get Nut working on a system with Fedora 8 i86_64
> > installed and an APC Smart-UPS USB.
> >
> > The 2.2.0 rpm does not seem to like the USB on this UPS.
>
> Can you elaborate?
See above, will do.

>
> Where did this RPM come from?
Fedora repositories via KYum including updates.

>
> What error messages did you get?
See above.

>
> > I have tried compiling the 2.2.2 version,
> > from source, but am getting warnings. I have tried building it on my i386
> > system (f8 as well) and I get no warnings also the 2.2.0 works fine on
> > that one with a TrippLite Omni 900 LCD via USB.
>
> I haven't verified all of the warnings you listed, but they look like
> they mostly come from debug messages, and probably will not cause any
> change during operation.
Thanks for the re-assurance, as I am not into development and program Pascal rather than C/C++ I was 
not sure how important these were.

>
> That's not to say that there aren't other places where the code isn't
> 64-bit clean, but given that NUT hardly ever handles anything large
> enough to overflow a 32-bit register, I'd say it is more productive to
> look at runtime error messages rather than compile-time warnings.
Again, will come back with details once I can get back on line.