<
fedora@drussell.dnsalias.com> wrote:
>
>
> On Fri, Sep 26, 2008 at 11:31 PM, Kam Leo <
kam.leo@gmail.com> wrote:
>>
>> On Fri, Sep 26, 2008 at 10:58 PM, Don Russell
>> <
fedora@drussell.dnsalias.com> wrote:
>> >
>> >
>> > On Thu, Sep 25, 2008 at 4:56 AM, Bill Crawford
>> > <
billcrawford1970@gmail.com>
>> > wrote:
>> >>
>> >> On Thursday 25 September 2008 12:41:13 Brian Millett wrote:
>> >>
>> >> > So the password field has changed from x to * ????
>> >> >
>> >> > I know that that means look in /etc/shadow for the password, but what
>> >> > inconsistancy will the older, established users find ??
>> >>
>> >> 'x' means look in /etc/shadow, '*' is one of several ways of indicating
>> >> "no
>> >> password" as in you can't log in, rather than "blank password" which
>> >> lets
>> >> all
>> >> log in without one. The .rpmnew is the "unconverted" form, if you run
>> >> pwunconv
>> >> you'll see the same it /etc/passwd.
>> >>
>> >
>> >
>> > The /etc/passwd.rpmnew has 15 lines of userid stuff...
>> > My /etc/passwd file has a lot more than that... and many I didn't even
>> > know
>> > about.... (various system things ntpd blah blah blah)
>> >
>> > Am I supposed to take the users that *I* added to the system (via
>> > system-config-users) and cut/paste those ones into the new one, changing
>> > the
>> > x to an *? And thereby dropping all those other ones that are set to
>> > nologon
>> > anyway?
>> >
>> > Does pwconv or pwunconv do this for me automatically? (The man file
>> > looks
>> > great for people familiar with it... not so great for explaining what
>> > the
>> > commands really do.) Shouldn't the update script have done this when it
>> > updated setup?
>> >
>> > The pwconv command creates shadow from passwd and an optionally
>> > existing
>> > shadow.
>> >
>> > The pwunconv command creates passwd from passwd and shadow and
>> > then
>> > removes shadow.
>> >
>> > So where does passwd.rpmnew come into play?
>> >
>> > pwconv ... and removes shadow... um, don't I need shadow?
>> > ditto for pwunconv
>> >
>> > I don't get it, now I don't know what I have. :-(
>> >
>>
>> The passwd rpm specified the creation of /etc/passwd. Since your
>> system already had an existing /etc/passwd file that spec instructions
>> were written such that the old file was not overwritten. Your old file
>> was protected and /etc/passwd.rpmnew was created. You should thank the
>> packager for being diligent.
>
> Yes, I understand that part of it. What I don't understand now is, what am I
> supposed to do with passwd.rpmnew?
>
> I don't mind manually merging in changes etc.. I do this all the time with
> other config files that get created as .rpmnew or .rpmsave... I always go
> and see what's changed....When I see this sort of thing for other packages
> (sendmail for example) I compare the two config files and it's pretty
> obvious they've added a new config option or something, but since I've
> changed mine, they leave the new one as rpmnew... no problema, I merge my
> changes in, restart sendmail, and presto keeno.. back in business with an
> up-to-date system.
>
> Right now, I'm left wondering what I'm supposed to do with these passwd
> files. Since I didn't add all those items to passwd in the first place, and
> now they are removed from passwd.rpmnew, does that mean those IDs are no
> longer required, and they should be removed from my passwd file?
>
> Every ID in passwd.rpmnew is already in passwd. But all IDs in passwd are
> not in passwd.rpmnew.
>
> As I asked before, am I supposed to drop those ones that were system
> generated, and just now keep the ones in passwd.rpmnew + the specific users
> *I* created?
>
> Should I just change the "x" in passwd to * and call it a day?
> And if that's the case, surely the update process could have done that...
> but, it's not a big deal.
>
> As for "thanking the packager for being diligent". Um, doing that sort of
> "save the file stuff" is so basic, it pretty much goes with saying. But,
> yes, I DO appreciate the efforts of everybody that contributes to Fedora
> (and other open source stuff)... including myself.
>
> To me, being "diligent" in this regard is making it obvious what needs to be
> done next if the script didn't "do it all".