[Smartmontools-support]Re: Some mods to proposed short/long options

[Bruce A. <ba...@gr...>]

Hi Erik,

Thanks for the quick reply!

On Mon, 2 Dec 2002, [ISO-8859-1] Erik Inge Bols=F8 wrote:

> On Mon, 2 Dec 2002, Bruce Allen wrote:
> >Thanks for making the table of proposed short/long options at the end of
> >the README file, and for setting up the source and Makefile with a
> >HAVE_GETOPT_LONG
> >#define so that long options can be turned off trivially.  Erik, I hope
> >you are OK with this solution. I think it makes it trivial to build
> >smartmontools with only short options, if desired or needed.
>=20
> By all means, as long as both sets of options and compile with/without ar=
e
> tested :)

Both Phil and I will see to this.

Peter, I think that this resolves at least one question: we WILL support
long options, provided HAVE_GETOPT_LONG is defined.  So for you the issue
is now (I think) "can I link to the standard glibc to resolve
getopt_long()".  Hopefully the answer is "yes" so that Darwin can do long
options without our having to include the getopt_long code with the
package.  At some point, please write and let us know about this...

> >** I would prefer the arguments to --tolerance to be strings, eg
> >   "permissive" and "conservative" rather than numerical.  Should
> >   we keep the third agument "normal" or drop it since it is the default=
?
> >   If there is no argument in favor of keeping it, Occam's Razor says to
> >   drop it...
>=20
> For consistency, I think you should be able to explicitly say you want
> "normal", even if it is the default.

Noted.

I am still not 100% convinced but weakening fast...

Bruce