Source Mage code commit list

[Jeremy Blosser <jblosser-smgl AT firinn.org>]

On Feb 06, David Kowis [dkowis AT shlrm.org] wrote:
> Jeremy Blosser wrote:
> > I haven't tested this anywhere yet... if we think this is worth changing
> > I'll do that then revert the other and apply this version.  (Including 
> > your
> > other changes to the other files as well, of course.)
> > 
> > diff --git a/security/shadow/CONFIGURE b/security/shadow/CONFIGURE
> > index 6c97b30..c06410e 100755
> > --- a/security/shadow/CONFIGURE
> > +++ b/security/shadow/CONFIGURE
> > @@ -1,5 +1,14 @@
> > -config_query SHADOW_CONV "Attempt to convert/fix accounts (with 
> > installed shadow utils)" n &&
> > -if [[ "$SHADOW_CONV" == y ]]; then
> > +if [[ "$SHADOW_CONV" == n ]] && [[ -z "$SHADOW_RECONV" ]]; then
> > +  message "Checking passwd for shadowness (there could have been an 
> > unwanted run of pwunconv/grpunconv, see bug #8834)." &&
> > +  if ! grep -q '^root:x:' "$INSTALL_ROOT/etc/passwd"; then
> > +    message "${PROBLEM_COLOR}Your passwd file contains password hashes, 
> > resuggesting conversion to shadow." &&
> > +    message "${MESSAGE_COLOR}You can still say n to the upcoming query 
> > but make sure to run pwconv / grpconv yourself if you want existing 
> > passwords shadowed." &&
> > +    query SHADOW_RECONV "Attempt to convert/fix accounts (with installed 
> > shadow utils)" n &&
> 
> Don't you want that n to be y? Since you're kinda suggesting that they
> do this?

IMO, no.  The point to me is to make the question get re-asked (instead of
taking the old answer and not asking).  But we should still assume n, they
only got here if they said n last time, and maybe they meant that.

Attachment: pgpcEwgOaMm89.pgp
Description: PGP signature