Subject: Re: making an existing feature optional (Re: CVS commit:
To: None <email@example.com>
From: YAMAMOTO Takashi <firstname.lastname@example.org>
Date: 08/07/2005 00:10:46
> > - introduce "default-options".
> > (an alternative might be just to allow "options" in one_def.)
> I don't like this. I think that the complete separation between the
> descriptive step and the operational step of the configuration is the
> one thing that makes sense in config(1).
config(1) already allows "makeoptions" in both steps.
and IMO default values belongs to the descriptive step.
> and leave room for other stuff (like, say, INET).
do you mean making INET default?
i don't think we can flip default without breaking existing config files.
> Yes, people who write their config files down to the "machine ..." line
> will have to edit their config file, but I'm rather confident those read
> current-users and UPDATING.
i'd say such people should read source-changes@. :)