Re: [patch 1/3] fix buggy IEEE80211_CRYPT_* selects

[Jeff Garzik <jgarzik@xxxxxxxxx>]

Adrian Bunk wrote:
> On Sun, Mar 06, 2005 at 01:00:50PM -0500, Jeff Garzik wrote:
>
> > Herbert Xu wrote:
> >
> > > On Sat, Mar 05, 2005 at 12:07:18AM +0100, Adrian Bunk wrote:
> > >
> > >
> > > > The way kconfig currently works, you have to ensure that the 
> > > > dependencies of what you select are fulfilled.
> > >
> > > Yes Adrian's right.  In the other places where we select CRYPTO
> > > symbols, we always make sure that CRYPTO itself is selected.  See
> > > net/ipv4/Kconfig for example.
> > I would rather fix Kconfig.  If we are selecting X_1 -- which explicitly 
> > depends on X -- when Kconfig should automatically select X.
> > It is completely illogical to duplicate a dependency chain each time you 
> > wish to select a symbol.
>
> I asked about the first example of my last email:
>   What values of the variables A-E do you expect exactly if the user 
>   turns on F?
> If you expect this to work, which unambiguous solution do you propose 
> for this example?
If that information is missing, it should re-prompt the user, as it does 
for a few other cases.
        Jeff