[Nut-upsdev] zero-prepended values - why oh why?
aquette.dev at gmail.com
Mon Feb 27 20:52:08 UTC 2006
2006/2/23, Niklas Edmundsson <nikke at acc.umu.se>:
> On Thu, 23 Feb 2006, Niels Baggesen wrote:
> > On Thu, Feb 23, 2006 at 02:24:49PM -0400, Peter Selinger wrote:
> >> Could you give some examples of what you mean? -- Peter
> > Try running "grep '%0[0-9]*\.[0-9]f' drivers/*.c" ... you will find
> > examples in 39 different files!
> Yup. snmp-ups is one of those culprits, mge-utalk is another.
and mge-shut too.
I should have cleaned it for long... now it's done.
I've done mge-utalk / mge-shut, and snmp-ups. For the latter, we'll
however have to rework the *mib files, and the driver core à la
For the others, can you contact the authors to check if it's needed.
If not, then the drivers will have to be cleaned.
> And regarding the statement that certain clients might depend on the
> width of the string: If that's the case then we're in trouble, because
> very few drivers are identical in behaviour ;)
I don't know any client that would do such a thing.
and they would be mad to do so
> > It is for instance the native number format for upscode2 devices.
> Yup, but there it actually has a reason for outputting what it does.
what is that reason?
More information about the Nut-upsdev